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
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.08979591836734693, 'partial': 0.017699115044247787}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.44, 'partial': 0.6666666666666666}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.14915254237288134, 'partial': 0.034482758620689655}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.04794520547945205, 'recall': 0.04794520547945205, 'f1': 0.04794520547945205, 'support': None}, 'macro_avg': {'precision': 0.021499006682318945, 'recall': 0.22133333333333333, 'f1': 0.036727060198714204, 'support': None}, 'weighted_avg': {'precision': 0.007960767044119138, 'recall': 0.04794520547945205, 'f1': 0.01330132182573677, 'support': None}}
Loss: tensor(0.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, 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
				 {'precision': {'support': 0.8873239436619719, 'agreement': 0.5166666666666667, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.42, 'partial': 0.0}, 'recall': {'support': 0.8873239436619719, 'agreement': 0.5904761904761905, 'direct_attack': 0.046153846153846156, 'undercutter_attack': 0.84, 'partial': 0.0}, 'f1': {'support': 0.8873239436619719, 'agreement': 0.5511111111111111, 'direct_attack': 0.08108108108108109, 'undercutter_attack': 0.5599999999999999, 'partial': 0.0}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7226027397260274, 'recall': 0.7226027397260274, 'f1': 0.7226027397260274, 'support': None}, 'macro_avg': {'precision': 0.4314647887323944, 'recall': 0.4727907960584017, 'f1': 0.41590322717083283, 'support': None}, 'weighted_avg': {'precision': 0.7053367579908676, 'recall': 0.7226027397260274, 'f1': 0.6954399605084537, 'support': None}}
Loss: tensor(1.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7799, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8085, 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
				 {'precision': {'support': 0.8821989528795812, 'agreement': 0.5555555555555556, 'direct_attack': 0.5, 'undercutter_attack': 0.40816326530612246, 'partial': 0.0}, 'recall': {'support': 0.9492957746478873, 'agreement': 0.5238095238095238, 'direct_attack': 0.03076923076923077, 'undercutter_attack': 0.8, 'partial': 0.0}, 'f1': {'support': 0.9145183175033921, 'agreement': 0.5392156862745098, 'direct_attack': 0.057971014492753624, 'undercutter_attack': 0.5405405405405406, 'partial': 0.0}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7431506849315068, 'recall': 0.7431506849315068, 'f1': 0.7431506849315068, 'support': None}, 'macro_avg': {'precision': 0.4691835547482518, 'recall': 0.4607749058453283, 'f1': 0.4104491117622392, 'support': None}, 'weighted_avg': {'precision': 0.7267502138200185, 'recall': 0.7431506849315068, 'f1': 0.7055938231876434, 'support': None}}
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.9018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683, 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(1.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8050, 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(1.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7154, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, 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
				 {'precision': {'support': 0.8910081743869209, 'agreement': 0.5643564356435643, 'direct_attack': 0.36363636363636365, 'undercutter_attack': 0.42, 'partial': 0.09090909090909091}, 'recall': {'support': 0.9211267605633803, 'agreement': 0.5428571428571428, 'direct_attack': 0.24615384615384617, 'undercutter_attack': 0.42, 'partial': 0.2222222222222222}, 'f1': {'support': 0.9058171745152355, 'agreement': 0.5533980582524272, 'direct_attack': 0.29357798165137616, 'undercutter_attack': 0.41999999999999993, 'partial': 0.1290322580645161}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7243150684931506, 'recall': 0.7243150684931506, 'f1': 0.7243150684931506, 'support': None}, 'macro_avg': {'precision': 0.4659820129151879, 'recall': 0.4704719943593183, 'f1': 0.4603650944967109, 'support': None}, 'weighted_avg': {'precision': 0.7209244402473914, 'recall': 0.7243150684931506, 'f1': 0.7207461510262562, 'support': None}}
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2124, 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(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(1.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, 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(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, 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(0.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8164, 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(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7658, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6687, 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(1.3747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4726, 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(1.7996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7907, 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(0.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(1.9656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, 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(0.8986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9314, 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(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4125, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, 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(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.2960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, 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(1.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(3.1197, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, 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(0.8231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1606, 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(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, 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(1.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9529, 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(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2276, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, 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(1.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5435, 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(0.9173, 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(0.0036, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7394, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5550, 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(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, 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(0.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0482, 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(0.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(1.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6157, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8463, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2771, 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(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2881, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5728, 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
				 {'precision': {'support': 0.8704663212435233, 'agreement': 0.6666666666666666, 'direct_attack': 0.2727272727272727, 'undercutter_attack': 0.4017857142857143, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9464788732394366, 'agreement': 0.38095238095238093, 'direct_attack': 0.046153846153846156, 'undercutter_attack': 0.9, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9068825910931173, 'agreement': 0.4848484848484849, 'direct_attack': 0.07894736842105263, 'undercutter_attack': 0.5555555555555556, 'partial': 0.4166666666666667}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7345890410958904, 'recall': 0.7345890410958904, 'f1': 0.7345890410958903, 'support': None}, 'macro_avg': {'precision': 0.5089958616513021, 'recall': 0.5658281311802439, 'f1': 0.48858013331697536, 'support': None}, 'weighted_avg': {'precision': 0.718890586443509, 'recall': 0.7345890410958904, 'f1': 0.7012187799183113, 'support': None}}
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4989, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3948, 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(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5542, 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(1.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5968, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4244, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2332, 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(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8440, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4328, 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(2.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8961, 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(2.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7729, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, 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(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5552, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4733, 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(0.0025, 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(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, 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(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7377, 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(2.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, 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(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3191, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5037, 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(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, 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(0.8922, 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(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(1.8286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6429, 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(0.9713, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, 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(0.4997, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4981, 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(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3618, 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(0.4899, 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(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, 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(0.2213, 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(0.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4970, 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(0.1781, 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(0.2001, 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(1.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, 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(0.0324, 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(1.8467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0731, 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(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7686, 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(0.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6908, 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(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, 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(6.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7538, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3450, 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(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4316, 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(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, 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(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8156, 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(0.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5130, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7642, 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(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5915, 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
				 {'precision': {'support': 0.9171779141104295, 'agreement': 0.460431654676259, 'direct_attack': 0.3877551020408163, 'undercutter_attack': 0.4883720930232558, 'partial': 0.14814814814814814}, 'recall': {'support': 0.8422535211267606, 'agreement': 0.6095238095238096, 'direct_attack': 0.2923076923076923, 'undercutter_attack': 0.42, 'partial': 0.4444444444444444}, 'f1': {'support': 0.8781204111600588, 'agreement': 0.5245901639344263, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.45161290322580644, 'partial': 0.2222222222222222}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.696917808219178, 'recall': 0.696917808219178, 'f1': 0.696917808219178, 'support': None}, 'macro_avg': {'precision': 0.48037698239978166, 'recall': 0.5217058934805413, 'f1': 0.48197580677516944, 'support': None}, 'weighted_avg': {'precision': 0.7275676418961624, 'recall': 0.696917808219178, 'f1': 0.7072979880186517, 'support': None}}
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, 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(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, 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(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0303, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, 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(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7390, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, 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(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, 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(1.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3122, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, 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(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, 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(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, 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(0.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, 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(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8186, 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(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, 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(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, 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(0.0004, 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(1.0899, 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(0.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, 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(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, 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(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3815, 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(0.1453, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0005, 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(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(0.0019, 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(0.0081, 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(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, 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(1.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6208, 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(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, 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(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0144, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, 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(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8410, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, 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(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, 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(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(2.9211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, 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(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7620, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, 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(0.0006, 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(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, 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(1.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3433, 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(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3180, 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(0.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, 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(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6812, 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(0.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, 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(0.0250, 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(0.2914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3578, 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(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, 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(0.0056, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, 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(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7009, 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(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1332, device='cuda:0', grad_fn=<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(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, 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(0.8304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7815, 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(0.2354, 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(0.0079, 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(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, 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(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, 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(0.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748, 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(2.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, 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(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, 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(0.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(3.4006, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, 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(1.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, 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(0.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, 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(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, 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(1.4413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, 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(0.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2569, 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(0.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, 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(1.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, 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(0.0071, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7918, 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(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, 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(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.3639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3951, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2781, 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(0.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8264, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, 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(0.0009, 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(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015, 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(0.5609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, 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(0.5363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4891, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7961, 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(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, 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(1.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0330, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, 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(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, 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(0.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6848, 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(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5751, 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(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, 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(0.5868, 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(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, 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(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, 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(0.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, 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(0.1570, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, 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(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, 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(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9198, 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(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1681, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6967, 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(0.0018, 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(0.0113, 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(4.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3191, 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(0.0959, 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(0.0002, 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(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, 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(1.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, 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(0.0174, 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(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, 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(1.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, 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(0.1992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2411, 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(0.0233, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, 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(0.2341, 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(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, 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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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
				 {'precision': {'support': 0.8716577540106952, 'agreement': 0.6666666666666666, 'direct_attack': 0.23529411764705882, 'undercutter_attack': 0.3191489361702128, 'partial': 0.15789473684210525}, 'recall': {'support': 0.9183098591549296, 'agreement': 0.4, 'direct_attack': 0.12307692307692308, 'undercutter_attack': 0.6, 'partial': 0.3333333333333333}, 'f1': {'support': 0.8943758573388204, 'agreement': 0.5, 'direct_attack': 0.16161616161616163, 'undercutter_attack': 0.4166666666666667, 'partial': 0.21428571428571427}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7003424657534246, 'recall': 0.7003424657534246, 'f1': 0.7003424657534246, 'support': None}, 'macro_avg': {'precision': 0.4501324422673478, 'recall': 0.47494402311303724, 'f1': 0.43738887998147263, 'support': None}, 'weighted_avg': {'precision': 0.705669725618057, 'recall': 0.7003424657534246, 'f1': 0.6905314805175283, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0651, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, 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(0.0058, 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(0.0126, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, 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(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.0071, 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(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, 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(1.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, 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(0.0196, 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(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7340, 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(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8268, 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(0.0766, 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(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2523, 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(0.0037, 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(0.0018, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7526, 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(0.0030, 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(0.0060, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4856, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, 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(2.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9126, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, 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(0.0636, 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(0.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, 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(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7043, 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(0.0003, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, 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(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, 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(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, 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(0.0044, 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(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3683, 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(0.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, 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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, 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(0.0130, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, 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(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, 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(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9047, 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(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, 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(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, 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(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, 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(0.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2523, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0007, 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(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, 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(0.5186, 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(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, 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(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, 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(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1933, 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(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4610, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0795, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, 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(1.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0084, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0015, 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(0.0012, 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(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, 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(0.0053, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, 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(0.0024, 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(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, 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(0.0005, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, 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(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0720, 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(1.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8935, 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(0.0003, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, 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(1.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2654, 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(0.0451, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(0.0243, 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(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, 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(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, 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(0.6369, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3691, 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(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8702, 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(1.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0466, 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(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4120, 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(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, 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(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4303, 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(0.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, 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(0.0113, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, 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(0.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, 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(0.7345, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(2.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, 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(0.0576, 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(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, 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(0.0033, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, 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(2.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3334, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, 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(0.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, 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(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, 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(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5840, 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(2.7871, 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(0.0384, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5686, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, 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(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, 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(0.0008, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0108, 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(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, 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(0.0026, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, 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(2.2502, device='cuda:0', grad_fn=<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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7773, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, 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(0.0155, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3352, 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(0.1136, 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(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(0.1680, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, 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(0.0084, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, 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(0.0669, device='cuda:0', grad_fn=<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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, 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(0.0111, 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(0.9015, 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(0.0145, 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(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0502, 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(2.4474, 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(0.0005, 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(1.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5680, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, 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(0.9257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3376, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, 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(0.0009, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, 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(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, 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(0.0197, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, 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(0.0007, 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(0.0602, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0006, 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(0.0035, 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(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, 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(5.9962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, 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(0.0030, 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(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, 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(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, 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(0.1360, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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
				 {'precision': {'support': 0.8842975206611571, 'agreement': 0.5813953488372093, 'direct_attack': 0.2777777777777778, 'undercutter_attack': 0.36893203883495146, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9042253521126761, 'agreement': 0.47619047619047616, 'direct_attack': 0.07692307692307693, 'undercutter_attack': 0.76, 'partial': 0.2222222222222222}, 'f1': {'support': 0.8941504178272981, 'agreement': 0.5235602094240838, 'direct_attack': 0.12048192771084337, 'undercutter_attack': 0.4967320261437908, 'partial': 0.17391304347826086}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7123287671232876, 'recall': 0.7123287671232876, 'f1': 0.7123287671232877, 'support': None}, 'macro_avg': {'precision': 0.45105196579364765, 'recall': 0.48791222548969027, 'f1': 0.44176752491685534, 'support': None}, 'weighted_avg': {'precision': 0.7067808274754026, 'recall': 0.7123287671232876, 'f1': 0.6962848704073944, 'support': None}}
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.1411, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, 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(0.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0006, 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(0.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, 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(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3713, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0066, 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(0.0057, 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(0.0006, 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(0.0409, 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(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, 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(0.0202, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, 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(0.0055, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_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(0.0033, 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(0.1032, 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(0.0116, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, 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(0.0622, 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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, 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(0.0006, 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(0.0030, 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(0.0069, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2071, 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(0.0002, 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(0.0676, 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(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, 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(0.0090, 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(0.0023, 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(2.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_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(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, 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(0.0036, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3841, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3892, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1874, 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(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0560, 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(0.0003, 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(0.0400, 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(0.0061, 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(0.0094, 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(0.2467, 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(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4844, 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(0.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2945, 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(1.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0003, 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(0.0002, 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(0.0226, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0065, 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(0.1002, 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(0.0006, 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(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0321, 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(0.0039, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(0.4914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9370, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, 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(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5569, device='cuda:0', grad_fn=<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(0.0011, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5399, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, 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(0.5918, 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(0.0006, 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(0.0250, 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(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, 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(1.2000, device='cuda:0', grad_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(0.0122, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, 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(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, 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(0.0792, 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(0.0181, device='cuda:0', grad_fn=<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(0.0324, 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(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0112, 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(0.0058, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4661, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, 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(0.0024, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0003, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, 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(0.0026, 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(0.0148, device='cuda:0', grad_fn=<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(0.0201, 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(0.0232, 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(0.0018, 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(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8289, 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(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, 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(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.6141, 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(0.0020, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6265, 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(0.0036, 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(0.0107, 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(0.0065, 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(0.0084, 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(0.0341, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0344, 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(0.0952, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8857, 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(0.0014, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2693, 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(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, 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(0.0024, 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(0.1497, 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(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5747, 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(0.0565, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(2.2626, 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(0.0002, 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(0.1630, 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(0.2110, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, 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(0.0001, 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(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, 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(0.0010, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0115, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, 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(0.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3697, 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(2.4265, 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(0.0065, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, 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(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4431, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, 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(0.0100, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, 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(0.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, 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(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2986, 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(1.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, 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(0.0820, 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(0.0150, 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(0.0228, 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(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, 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(1.8838, 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(0.1792, 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(0.0202, 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(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, 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(0.1832, 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(1.7490, 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(0.3994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4420, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, 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(0.0544, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, 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(0.0125, 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(0.0010, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1828, 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(0.0011, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.8982, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8753, 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(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, 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(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3022, 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(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3184, 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(0.0006, 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(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3675, 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(0.0200, 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(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, 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(0.4703, 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(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3095, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5311, 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(0.4412, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, 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(0.2926, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0296, 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(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8548, 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(0.0067, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0163, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<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(0.0092, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(0.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0073, 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(0.2221, 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(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2157, 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(0.0003, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, 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(0.0002, 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(0.0012, 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(0.0002, 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(0.0012, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3017, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8907103825136612, 'agreement': 0.6619718309859155, 'direct_attack': 0.3125, 'undercutter_attack': 0.37037037037037035, 'partial': 0.11764705882352941}, 'recall': {'support': 0.9183098591549296, 'agreement': 0.44761904761904764, 'direct_attack': 0.15384615384615385, 'undercutter_attack': 0.6, 'partial': 0.4444444444444444}, 'f1': {'support': 0.9042995839112343, 'agreement': 0.5340909090909092, 'direct_attack': 0.2061855670103093, 'undercutter_attack': 0.4580152671755725, 'partial': 0.18604651162790695}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.714041095890411, 'recall': 0.714041095890411, 'f1': 0.7140410958904111, 'support': None}, 'macro_avg': {'precision': 0.47063992853869524, 'recall': 0.5128439010129151, 'f1': 0.45772756776318646, 'support': None}, 'weighted_avg': {'precision': 0.7287655309825363, 'recall': 0.714041095890411, 'f1': 0.7107588040447492, 'support': None}}
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.1958, 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(0.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, 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(0.1512, device='cuda:0', grad_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(0.0044, 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(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, 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(0.0100, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, 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(0.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, 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(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, 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(0.0254, 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(0.0008, 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(0.0803, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0009, 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(0.0008, 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(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6782, 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(0.0666, 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(0.0143, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7519e-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(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, 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(0.0096, device='cuda:0', grad_fn=<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(0.0118, 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(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, 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(0.0032, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, 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(2.6896, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<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(0.0006, 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(0.0013, 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(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3070, 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(1.2317, 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(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0093, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, 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(0.0034, 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(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, 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(0.8819, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4643, device='cuda:0', grad_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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2652, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7473, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0268, 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(0.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, 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(0.0006, 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(0.3713, 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(0.0001, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0107, 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(0.0033, 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(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3599, 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(0.0137, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, 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(0.3181, 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(0.0617, device='cuda:0', grad_fn=<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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0002, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, 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(0.0003, 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(0.1429, 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(0.0010, 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(0.0001, 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(0.0131, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, 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(0.0002, 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(0.0013, 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(0.0026, device='cuda:0', grad_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(0.0018, 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(0.0937, 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(0.0044, 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(0.0026, 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(0.0022, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922, 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(0.0022, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, 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(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, 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(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2850, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, 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(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, 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(0.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, 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(0.0045, 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(0.0004, 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(0.4694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, 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(0.0026, 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(0.0090, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2907, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2931, 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(0.0051, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0303, 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(0.0063, 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(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7004e-05, 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(8.9073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, 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(0.0016, 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(0.0058, 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(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, 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(0.0990, 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(0.0001, 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(0.2178, 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(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, 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(0.0058, 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(0.0002, 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(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3098, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(4.8424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, 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(0.0023, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, 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(0.0119, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, 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(0.0014, 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(0.0011, 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(0.0112, 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(0.0152, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, 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(0.1099, 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(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, 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(0.0062, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0360, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, 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(0.3938, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6076, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9970e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, 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(0.0002, 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(0.2279, 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(0.0002, 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(0.0099, 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(0.0107, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0289, 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(7.0212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<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(0.0908, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, 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(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.0825, 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(0.0011, 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(0.0072, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0050, device='cuda:0', grad_fn=<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(0.8837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, 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(0.0077, 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(0.0136, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7600, 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(8.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0007, 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(0.0683, 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(0.0019, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0074, 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(0.1547, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, 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(0.0018, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716, 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(0.0318, 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(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, 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(0.1796, 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(0.4987, 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(0.0008, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-05, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-05, 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(0.0002, 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(0.0002, 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(0.0023, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<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(0.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7709e-05, 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(0.1470, 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(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0009, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, 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(0.0004, 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(0.0224, 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(0.0065, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0303e-05, 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(0.3780, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, 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(0.0001, 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(0.1187, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6124, 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(0.9506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, 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(0.0007, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, 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(7.5177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0017, 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(0.0038, 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(0.1611, 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(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7679, 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(7.5965e-05, 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(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<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(0.0023, 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(1.8533, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1980e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7086, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4315, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, 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(0.0004, 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(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(9.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, 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(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, 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(0.0003, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, 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(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(0.0629, 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(0.0399, 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(0.0122, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, 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(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2915, 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(0.0271, 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(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0009, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3990, 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(0.0267, 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(0.1540, 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(1.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4511e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2695, 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(0.0508, 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(0.0440, 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(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(1.3023, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, 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(0.0078, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0012, 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(8.1686e-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(2.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, 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(0.0187, 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
				 {'precision': {'support': 0.8779840848806366, 'agreement': 0.6282051282051282, 'direct_attack': 0.47368421052631576, 'undercutter_attack': 0.453125, 'partial': 0.08695652173913043}, 'recall': {'support': 0.9323943661971831, 'agreement': 0.4666666666666667, 'direct_attack': 0.13846153846153847, 'undercutter_attack': 0.58, 'partial': 0.4444444444444444}, 'f1': {'support': 0.9043715846994536, 'agreement': 0.53551912568306, 'direct_attack': 0.2142857142857143, 'undercutter_attack': 0.5087719298245614, 'partial': 0.14545454545454548}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7226027397260274, 'recall': 0.7226027397260274, 'f1': 0.7226027397260274, 'support': None}, 'macro_avg': {'precision': 0.5039909890702422, 'recall': 0.5123934031539665, 'f1': 0.4616805799894669, 'support': None}, 'weighted_avg': {'precision': 0.7395106523527862, 'recall': 0.7226027397260274, 'f1': 0.7156809582087633, 'support': None}}
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, 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(0.0031, 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(0.0877, 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(0.0147, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8922e-05, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, 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(0.1529, 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(0.0006, 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(0.0002, 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(0.0085, 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(0.0007, 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(0.0159, 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(0.2221, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3745, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, 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(0.0005, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_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(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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8689e-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(0.0029, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5852, 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(0.0013, 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(0.0003, 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(0.1039, 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(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0038, 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(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3432e-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(0.0672, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0026, 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(0.0018, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0179, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8479, 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(0.0026, 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(0.0070, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1442, 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(1.1479, 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(6.7396e-05, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0208, 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(0.0151, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0079, 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(0.0101, 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(0.0017, 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(0.0394, 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(0.0008, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6090, 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(0.0005, 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(0.0031, 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(0.0069, 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(0.0144, 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(0.0014, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, 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(9.5189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.9501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5501e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, 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(0.0400, device='cuda:0', grad_fn=<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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2059e-05, 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(0.2465, 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(0.0493, 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(0.0056, 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(0.4905, device='cuda:0', grad_fn=<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(0.0172, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, 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(0.0004, 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(0.0037, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0032, 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(0.0183, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0017, 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(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8397e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, 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(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, 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(0.5486, 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(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, 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(0.0063, 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(0.0083, 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(7.9507e-05, 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(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, 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(0.0056, 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(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9423, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0096, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0547, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0023, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0198, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9607e-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(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0003, 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(0.0016, 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(0.0001, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, 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(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.0001, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, 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(0.0061, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8912e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6256, 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(0.0029, 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(0.0077, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1687e-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(0.0057, 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(0.0035, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-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(0.0328, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5964e-05, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, 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(0.0869, 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(0.0091, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8087, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0040, 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(0.0045, 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(0.1043, 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(0.0686, 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(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<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(0.0018, 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(0.0002, 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(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0008, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, 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(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0008, 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(6.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, 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(0.0005, 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(0.0024, 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(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-05, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3048e-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(0.0012, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1857e-05, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0332, 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(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, 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(9.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0100, 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(7.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0009, 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(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-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(0.0002, 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(0.0005, 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(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(0.0030, 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(0.0161, 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(0.0026, 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(6.7881e-05, device='cuda:0', grad_fn=<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(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9486e-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(0.0028, 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(0.0429, 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(3.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4055, 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(0.0018, 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(0.0034, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269, 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(0.0779, 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(0.0022, 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(8.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3797, 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(6.0948e-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(0.0013, 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(8.3895e-05, 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(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4470, 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(0.0006, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3157, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0007, 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(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8477, 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(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0995, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, 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(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, 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(3.9285, 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(0.0085, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5565, 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(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, 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(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4431, 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(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, 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(1.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, 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(0.0730, 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(0.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0798, 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(0.3111, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, 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(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, 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(0.7470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, 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(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1837, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2448, 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(0.0005, 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(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, 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(0.2476, 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(9.0860e-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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1225, device='cuda:0', grad_fn=<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(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3497, 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(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6089, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4286, 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(0.0543, 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(3.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3752, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3372, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8029, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, 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(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, 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(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, 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(0.0630, 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(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, 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(3.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, 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(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4448, 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(0.0682, 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(0.9170, 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(0.2390, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0877, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6249, 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(2.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, 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(0.0002, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.0012, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, 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(0.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, 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(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(0.0028, 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(0.1196, 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>)
			------------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
				 {'precision': {'support': 0.9002932551319648, 'agreement': 0.4931506849315068, 'direct_attack': 0.5, 'undercutter_attack': 0.5483870967741935, 'partial': 0.08695652173913043}, 'recall': {'support': 0.8647887323943662, 'agreement': 0.6857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.68, 'partial': 0.2222222222222222}, 'f1': {'support': 0.882183908045977, 'agreement': 0.5737051792828685, 'direct_attack': 0.15584415584415584, 'undercutter_attack': 0.6071428571428571, 'partial': 0.125}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7208904109589042, 'recall': 0.7208904109589042, 'f1': 0.7208904109589042, 'support': None}, 'macro_avg': {'precision': 0.5057575117153591, 'recall': 0.5090065865277134, 'f1': 0.46877522006317174, 'support': None}, 'weighted_avg': {'precision': 0.7398748133972904, 'recall': 0.7208904109589042, 'f1': 0.7106615482329384, 'support': None}}
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, 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(0.1820, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8395e-05, 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(1.8985, 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(0.0007, 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(1.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4457, 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(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8377e-05, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2388, 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(8.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2767, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2686e-05, 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(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0017, 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(0.0023, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, 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(0.4476, 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(1.2231, 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(9.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, 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(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_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(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(1.8316, 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(0.0116, 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(0.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2904, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8087, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8273, 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(0.0023, 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(0.0006, 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(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<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(0.0637, 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(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4897, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, 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(0.0035, 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(0.4981, 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(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0045, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5407, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, 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(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5005, 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(0.1367, 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(0.0011, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, 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(0.0474, 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(0.0007, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5545, 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(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, 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(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, 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(0.0104, 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(0.0002, 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(0.0098, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, 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(0.0050, 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(1.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, 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(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, 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(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8712, 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(0.0012, 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(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9203, 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(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.1306, 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(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.5830, 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(0.0068, 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(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, 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(0.0821, 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(0.0011, 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(0.2860, 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(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.1495, device='cuda:0', grad_fn=<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(0.0088, 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(0.0017, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0091, 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(0.0134, 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(0.0029, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, 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(0.1776, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, 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(0.0314, 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(0.0791, 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(0.0005, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3077, 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(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8549, 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(0.0070, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5141, 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(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, 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(0.0019, 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(0.0077, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.1271, device='cuda:0', grad_fn=<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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8860, 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(1.4394, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, 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(0.0072, 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(0.6571, 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(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0092, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, 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(0.0101, device='cuda:0', grad_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(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4599, 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(0.0016, 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(0.1028, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0070, 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(0.0221, 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(0.0006, 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(0.0104, 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(0.4188, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2094, 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(0.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, 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(0.0001, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5570, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, 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(0.1944, 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(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, 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(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, 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(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, 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(0.3403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, 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(3.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, 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(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, 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(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, 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(0.0035, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2119, 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(6.1754, 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(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8320, 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(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, 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(0.0039, 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(0.1390, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, 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(0.0480, 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(0.0011, 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(0.0003, 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(0.9483, 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(0.0233, 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(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, 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(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6678, 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(0.0167, 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(0.1239, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, 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(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, 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(0.0607, 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(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, 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(9.4280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5773e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5510, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-05, 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(0.0002, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, 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(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, 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(1.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3627, 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(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, 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(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2789, 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(0.0071, 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(0.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, 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(0.0004, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, 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(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, 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(0.4922, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, 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(0.9759, 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(0.5389, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, 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(6.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, 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(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3000, 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(0.0020, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, 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(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, 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(0.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, 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(0.0076, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3776, 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(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, 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(0.5410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0006, 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(0.0031, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6306, 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(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, 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(0.0435, 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(0.0159, 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(0.0018, 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(1.6328, 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(0.0027, 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(0.0058, 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(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0094, 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(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, 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(0.0347, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, 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(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, 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(0.0634, 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(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, 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(0.0009, 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(0.1547, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6406, 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(0.0002, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(4.2674, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(1.5531, 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(0.0014, 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(0.0373, 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(0.0873, 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(2.5038, device='cuda:0', grad_fn=<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(0.0044, 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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3962, 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(0.0835, 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(0.0009, 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(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9921e-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>)
			------------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
				 {'precision': {'support': 0.8972222222222223, 'agreement': 0.5714285714285714, 'direct_attack': 0.45, 'undercutter_attack': 0.39080459770114945, 'partial': 0.05263157894736842}, 'recall': {'support': 0.9098591549295775, 'agreement': 0.5333333333333333, 'direct_attack': 0.13846153846153847, 'undercutter_attack': 0.68, 'partial': 0.1111111111111111}, 'f1': {'support': 0.9034965034965036, 'agreement': 0.5517241379310344, 'direct_attack': 0.21176470588235297, 'undercutter_attack': 0.4963503649635037, 'partial': 0.07142857142857142}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7243150684931506, 'recall': 0.7243150684931506, 'f1': 0.7243150684931506, 'support': None}, 'macro_avg': {'precision': 0.47241739405986233, 'recall': 0.4745530275671121, 'f1': 0.4469528567403933, 'support': None}, 'weighted_avg': {'precision': 0.7324962379871107, 'recall': 0.7243150684931506, 'f1': 0.7155776960571963, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2529, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, 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(0.0567, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8186e-05, 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(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, 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(0.0003, 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(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<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(0.0042, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(1.3182, 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(0.0142, 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(0.0326, 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(0.0003, 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(0.0035, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, 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(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, 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(0.0042, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, 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(0.1435, 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(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7761e-05, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-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(0.4390, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0266, 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(0.0024, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, 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(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6074e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, 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(0.1140, 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(0.0525, 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(0.0016, 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(8.1898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8689e-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(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.2282, 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(0.0008, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0116, 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(7.6630e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, 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(0.0008, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, 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(0.0001, 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(0.0009, 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(0.0002, 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(0.0142, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<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(0.0018, 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(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0005, 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(0.0056, 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(0.0032, 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(0.0019, 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(0.0006, 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(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0396, 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(1.5825, 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(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0002, 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(0.0006, 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(0.0002, 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(0.0044, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, 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(0.0041, 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(0.0036, 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(0.0025, 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(0.0032, 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(0.0019, 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(0.0014, 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(0.0002, 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(0.0017, 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(0.0024, 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(0.0036, 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(0.8756, 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(0.0333, 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(0.0066, 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(0.0033, 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(0.1684, 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(0.0004, 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(0.0006, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, 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(0.0017, 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(0.0001, 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(0.0005, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, 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(0.0028, 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(0.0340, 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(0.0043, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0222e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5411, 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(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0145, 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(0.0054, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6749, 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(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, 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(0.0042, 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(0.1741, 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(0.0032, 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(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(8.1171e-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(0.0034, 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(0.0613, 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(0.0002, 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(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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4599, 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(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0066, 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(0.0027, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, 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(0.0098, device='cuda:0', grad_fn=<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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, 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(0.0297, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0313e-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(0.0080, 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(0.0032, 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(0.0162, 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(0.0001, 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(0.0578, 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(0.0019, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, 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(0.1280, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2966, 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(0.0366, 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(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0138, 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(0.0003, 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(0.0071, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4915, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, 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(0.0050, 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(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, 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(0.0017, 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(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8578, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3883, 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(0.0019, 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(0.0075, 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(0.5838, 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(0.0023, 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(7.7113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6275, 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(0.0105, 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(0.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0099, 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(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, 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(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9698, device='cuda:0', grad_fn=<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(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7904, 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(0.2259, 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(4.7233e-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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4753, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1524e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7722, 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(0.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2448, 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(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5612, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7133e-05, 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(0.0013, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0444, 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(0.0006, 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(0.2073, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, 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(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6799, 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(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1815, 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(0.0045, 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(0.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, 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(0.0047, 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(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, 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(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4525, 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(7.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9219, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-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(0.8400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2959, 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(0.4533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9221, 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(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2595e-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(0.0030, 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(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, 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(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6498, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-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(0.0705, 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(0.0403, 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(0.0004, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4593e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3117e-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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8975, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, 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(0.0041, 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(0.0134, 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(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4966e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, 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(0.0004, 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(0.0037, 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(0.6429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, 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(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<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(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, 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(6.9940e-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(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6762e-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(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, 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(0.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2078e-05, 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(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(1.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7742, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3732e-05, device='cuda:0', grad_fn=<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(0.0311, 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(0.0037, 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(0.0004, 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(0.0244, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, 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(0.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(0.0071, 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(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, 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(0.0291, 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(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4670, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2279, 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(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0619, 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(0.0021, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, 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(0.0024, 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
				 {'precision': {'support': 0.888283378746594, 'agreement': 0.5471698113207547, 'direct_attack': 0.4074074074074074, 'undercutter_attack': 0.43243243243243246, 'partial': 0.2}, 'recall': {'support': 0.9183098591549296, 'agreement': 0.5523809523809524, 'direct_attack': 0.16923076923076924, 'undercutter_attack': 0.64, 'partial': 0.2222222222222222}, 'f1': {'support': 0.9030470914127424, 'agreement': 0.5497630331753556, 'direct_attack': 0.2391304347826087, 'undercutter_attack': 0.5161290322580645, 'partial': 0.2105263157894737}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7345890410958904, 'recall': 0.7345890410958904, 'f1': 0.7345890410958903, 'support': None}, 'macro_avg': {'precision': 0.4950586059814378, 'recall': 0.5004287605977746, 'f1': 0.48371918148364895, 'support': None}, 'weighted_avg': {'precision': 0.7237954327856562, 'recall': 0.7345890410958904, 'f1': 0.7218347648130377, 'support': None}}
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2938, 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(0.0035, 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(1.6183, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, 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(0.0101, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0014, 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(0.0537, 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(0.0049, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, 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(0.0083, 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(0.0062, 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(0.0033, 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(0.0039, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, 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(0.0191, 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(0.0023, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4411, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, 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(0.0244, 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(0.5534, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, 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(0.0087, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0039, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, 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(0.0683, 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(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2374, 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(0.0117, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3856e-05, 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(0.0017, 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(0.0271, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0027, 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(0.0010, 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(1.5854, 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(5.5317e-05, 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(0.0013, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0002, 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(0.0015, 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(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6518, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, 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(0.0003, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, 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(1.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, 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(0.0054, 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(0.0003, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, 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(0.0057, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0013, 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(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, 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(0.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, 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(0.0116, 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(0.0034, 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(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3683, 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(0.0343, 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(0.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, 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(3.0103, 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(0.0012, 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(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, 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(0.0009, 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(0.0098, 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(0.0027, 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(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0029, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, 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(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, 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(0.0006, 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(0.0332, 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(0.1157, 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(0.0034, 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(0.2704, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<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(0.0006, 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(0.0140, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1437, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(0.0303, 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(0.0097, 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(0.4041, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, 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(0.0007, 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(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, 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(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, 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(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, 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(0.1967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, 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(0.0034, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0154, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, 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(0.0561, 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(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4345, 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(0.1346, 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(0.0041, 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(0.0005, 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(0.0122, 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(0.0059, 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(0.0061, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, 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(0.0499, 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(0.0003, 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(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, 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(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5190, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8084e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, 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(0.0042, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6368e-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(0.0003, 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(0.0008, 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(0.0004, 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(8.7196e-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(0.0020, 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(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9042e-05, 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(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1020, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(6.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0250, 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(0.0077, 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(0.0003, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0556e-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(0.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3443, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, 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(0.0050, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0887, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, 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(2.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0030, 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(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8542, device='cuda:0', grad_fn=<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(0.0002, 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(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0027, 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(0.0057, 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(0.0006, 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(0.0029, 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(8.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, 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(0.0028, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<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(0.4031, 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(0.0011, 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(0.0037, 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(0.0005, 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(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9454e-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(0.1390, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0058, 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(0.1683, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0061, 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(0.0109, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0031, 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(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.0005, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, 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(0.0003, 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(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, 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(0.0118, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-05, 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(0.0033, 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(0.0002, 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(0.0829, 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(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4281e-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(0.0484, 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(0.0232, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, 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(0.0079, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3235, 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(0.0032, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0363e-05, 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(0.0071, 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(8.6319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4965, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1527, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, 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(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, 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(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0002, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, 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(0.0023, 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(0.0071, 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(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0003, 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(0.0018, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0014, 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(0.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6631e-05, device='cuda:0', grad_fn=<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(0.6277, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8383e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-05, 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(0.1785, 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(0.0004, 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(0.5006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-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(0.0006, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5901, 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(0.0017, 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(0.0219, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0194, 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(3.5365e-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(0.0006, 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(0.0003, 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(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-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(0.1163, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2525e-05, 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(0.0134, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9765, 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(0.0010, 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(0.0020, 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(0.0071, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, 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(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773, 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(0.0003, 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(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0013, 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(0.0016, 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(0.0053, 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(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0020, 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(6.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1503e-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(0.0046, 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(0.0069, 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(0.0020, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, 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(8.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<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(0.0010, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8542713567839196, 'agreement': 0.6964285714285714, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.47058823529411764, 'partial': 0.1388888888888889}, 'recall': {'support': 0.9577464788732394, 'agreement': 0.37142857142857144, 'direct_attack': 0.046153846153846156, 'undercutter_attack': 0.8, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9030544488711819, 'agreement': 0.48447204968944096, 'direct_attack': 0.08108108108108109, 'undercutter_attack': 0.5925925925925927, 'partial': 0.22222222222222227}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7311643835616438, 'recall': 0.7311643835616438, 'f1': 0.7311643835616438, 'support': None}, 'macro_avg': {'precision': 0.49870207714576614, 'recall': 0.5461768904022426, 'f1': 0.45668447889130376, 'support': None}, 'weighted_avg': {'precision': 0.7240366611124384, 'recall': 0.7311643835616438, 'f1': 0.6992359494290424, 'support': None}}
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, 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(8.8074e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0073, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-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(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7448e-05, 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(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(0.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5280, 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(0.0040, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0001, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0121, 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(7.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, 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(0.0023, 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(0.0029, 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(4.1753e-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(0.0088, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3670, device='cuda:0', grad_fn=<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(0.0642, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, 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(0.0049, 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(0.0029, 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(0.0003, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1172e-05, 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(0.6499, 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(0.0001, 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(0.0043, 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(0.0684, 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(0.0026, 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(0.0006, 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(0.0086, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, 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(0.0106, 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(0.0159, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5488e-05, 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(0.0003, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, 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(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6610e-05, 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(0.0011, 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(0.0006, 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(0.0055, 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(8.5319e-05, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1493e-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(0.0025, 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(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-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(0.0006, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0155, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0821, 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(0.0013, 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(0.0008, 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(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, 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(6.4217e-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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(9.2063, 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(0.0010, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, 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(0.0021, 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(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8302, 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(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2405, 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(0.0056, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, 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(0.0016, 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(0.0007, 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(0.0002, 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(0.0019, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, 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(0.0915, 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(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4488e-05, device='cuda:0', grad_fn=<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(0.0048, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, 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(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-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(0.0243, 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(0.0043, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9194e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5669e-05, 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(8.2867e-05, 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(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-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(0.0016, 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(0.0131, 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(0.0007, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9083e-05, 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(0.0005, 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(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, 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(0.0235, 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(0.0019, 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(0.0005, 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(6.9727e-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(8.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_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(0.0011, 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(9.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, 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(0.0027, 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(0.0141, 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(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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, 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(9.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4530, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, 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(0.0029, 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(0.0014, 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(0.0035, 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(0.0041, 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(0.3217, 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(0.0016, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8104e-05, device='cuda:0', grad_fn=<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(0.0067, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0026, 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(0.0039, 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(0.0010, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, 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(0.0010, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6851e-05, device='cuda:0', grad_fn=<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(8.7862e-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(9.6577e-05, 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(0.0014, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-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(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(0.0020, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-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(8.5076e-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(0.0014, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-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(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7669e-05, 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(0.0009, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0347, 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(5.5044e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-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(0.0233, 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(0.0095, 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(0.0015, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0013, 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(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0046, 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(0.0006, 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(5.0533e-05, 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(0.0002, 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(0.0001, 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(0.0062, 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(9.5367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, 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(7.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, 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(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-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(0.0051, 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(0.0032, 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(0.0093, 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(0.0015, 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(0.0040, device='cuda:0', grad_fn=<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(0.0053, 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(0.0001, 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(0.0092, device='cuda:0', grad_fn=<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(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(9.4003e-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(0.0018, 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(0.0004, 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(0.0004, 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(0.0004, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-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(0.0062, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-05, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(6.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, 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(0.0008, 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(0.0007, 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(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, 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(0.0043, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-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(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, 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(3.3306e-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(7.9900e-05, 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(0.0154, 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(0.0003, 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(5.3409e-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(0.0009, 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(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, 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(0.0001, 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(0.0019, 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(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(7.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5357, 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(0.0013, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-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(0.0013, 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(2.3557e-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(6.1493e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, 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(4.3569e-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(3.3851e-05, 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(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9373e-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(0.0034, 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(6.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9486e-05, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1908e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-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(0.0010, 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(4.4357e-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(0.0008, 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(0.0096, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, 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(2.8522e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, 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(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-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(0.0110, 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(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, 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(0.0084, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-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(0.0032, 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(0.0067, 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(4.4387e-05, 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(8.0506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4689, 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(0.0014, 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(0.0002, 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(0.0032, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-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(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5895, 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(6.1462e-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(0.0073, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-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(0.0002, 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(0.0013, 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(0.0008, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, 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(5.3591e-05, 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(0.0016, 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(0.0229, 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(0.0005, 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(0.0093, 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(4.8020e-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(7.3662e-05, 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(6.3158e-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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-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(0.0018, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, 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(0.0020, 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(0.0069, 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(0.0001, 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(7.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0030, 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(0.0007, 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(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-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(0.0011, 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(0.0039, 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(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5843e-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(0.0005, 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(0.0020, 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(0.0010, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-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(4.8353e-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(0.0189, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0011, 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(0.0030, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6669e-05, device='cuda:0', grad_fn=<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(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3876e-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(0.0047, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, 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>)
			------------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
				 {'precision': {'support': 0.8786279683377308, 'agreement': 0.6666666666666666, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.43283582089552236, 'partial': 0.10416666666666667}, 'recall': {'support': 0.9380281690140845, 'agreement': 0.47619047619047616, 'direct_attack': 0.07692307692307693, 'undercutter_attack': 0.58, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9073569482288827, 'agreement': 0.5555555555555556, 'direct_attack': 0.125, 'undercutter_attack': 0.4957264957264957, 'partial': 0.1754385964912281}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7226027397260274, 'recall': 0.7226027397260274, 'f1': 0.7226027397260274, 'support': None}, 'macro_avg': {'precision': 0.48312609117998395, 'recall': 0.5253394555366386, 'f1': 0.45181551920043245, 'support': None}, 'weighted_avg': {'precision': 0.7297241206700981, 'recall': 0.7226027397260274, 'f1': 0.7105056885433776, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-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(7.5600e-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(0.0045, 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(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4237e-05, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6588e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-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(8.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0045, 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(0.0005, 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(0.0022, 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(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(0.0015, 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(0.0167, 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(0.0002, 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(0.0008, 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(9.1976e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, 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(0.0012, 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(0.0017, 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(0.0022, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-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(8.2926e-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(0.0031, 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(5.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-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(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7617e-05, device='cuda:0', grad_fn=<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(2.8159e-05, 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(0.0002, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-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(0.0002, 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(2.1558e-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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8727e-05, 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(0.0007, 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(0.0013, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4277e-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(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(0.0001, 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(0.0023, 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(0.0016, 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(0.0074, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-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(0.0008, 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(0.0016, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0009, 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(0.0001, 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(2.5434e-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(0.0009, 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(0.0001, 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(0.0043, 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(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0007, 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(2.0650e-05, device='cuda:0', grad_fn=<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(3.0793e-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(0.0015, 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(0.0001, 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(4.9352e-05, 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(3.6515e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, 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(2.3708e-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(0.0002, 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(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3644e-05, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-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(0.0267, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5701e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0960e-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(5.6195e-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(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, 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(0.0020, 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(9.6762e-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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2721e-05, 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(0.0003, 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(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-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(0.0006, 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(0.0020, 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(0.0001, 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(0.0007, 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(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(0.0004, 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(7.9020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, 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(0.0039, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-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(0.0007, 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(0.0020, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, 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(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-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(2.5343e-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(3.8453e-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(0.0011, 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(0.0004, 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(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-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(3.0460e-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(0.0012, 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(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7879e-05, 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(0.0022, 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(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, 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(2.8219e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6598e-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(0.0040, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9272e-05, 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(4.3024e-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(0.0003, 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(4.7536e-05, 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(0.0006, 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(0.0004, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, 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(4.9049e-05, 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(0.0057, 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(0.0040, 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(0.0004, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<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(4.8717e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5912e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9079e-05, 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(0.4506, 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(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<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(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3027e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2308e-05, 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(0.0011, 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(2.7190e-05, device='cuda:0', grad_fn=<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(4.3994e-05, 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(0.0002, 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(6.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0575e-05, 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(0.0008, 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(0.0003, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0566e-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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-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(0.0015, 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(0.0013, 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(7.2542e-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(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, 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(0.0015, 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(0.0010, 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(0.0009, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6116e-05, device='cuda:0', grad_fn=<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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-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(5.6498e-05, 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(0.0006, 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(0.0003, 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(0.0021, 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(0.0035, 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(0.0014, 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(7.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6032e-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(0.0003, 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(4.7929e-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(1.8379e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-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(0.0002, 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(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(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-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(9.9848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(2.5161e-05, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-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(0.0068, 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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-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(0.0022, 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(0.0012, 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(0.0131, device='cuda:0', grad_fn=<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(0.0010, 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(0.0005, 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(0.0006, 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(0.0010, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3490e-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(0.0006, 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(0.0036, 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(5.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5861e-05, 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(0.0009, 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(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-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(2.0832e-05, device='cuda:0', grad_fn=<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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2966e-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(0.0008, 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(0.0022, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-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(0.0002, 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(0.0004, 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(0.0023, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8041e-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(0.0015, 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(3.0642e-05, device='cuda:0', grad_fn=<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(3.9906e-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(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4459e-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(0.0014, 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(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-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(2.9885e-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(2.6978e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-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(7.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0855e-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(0.0013, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0004, 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(0.0004, 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(4.6961e-05, 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(0.0006, 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(0.0011, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-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(0.0003, 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(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0012, 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(0.0003, 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(6.3943e-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(0.0010, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3025e-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(6.3520e-05, device='cuda:0', grad_fn=<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(3.4668e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-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(2.6645e-05, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0015, 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(0.0015, 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(0.0026, device='cuda:0', grad_fn=<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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-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(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8627e-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(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-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(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8013e-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(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7081e-05, device='cuda:0', grad_fn=<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(3.1308e-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(9.3191e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-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(4.4841e-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(0.0007, 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(7.3330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2238e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(2.3102e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0473e-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(0.0013, 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(0.0016, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-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(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, 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(0.0005, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8947e-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(0.0169, 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(3.4911e-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(2.2406e-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(5.4954e-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(6.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4971e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-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(0.0005, 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(0.0062, 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(2.1679e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-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(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(4.1874e-05, 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(0.0014, 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(0.0006, 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(2.2891e-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(4.2298e-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(1.7410e-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(0.9045, 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(5.5922e-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(0.0012, 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(0.0016, device='cuda:0', grad_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(0.0007, 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(0.0004, 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(1.9257e-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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0017, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-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(4.0300e-05, device='cuda:0', grad_fn=<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(0.0025, 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(0.0047, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-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(2.5949e-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(0.0016, 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(7.1332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-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(2.4616e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7266e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0009, 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(0.0021, 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(0.0008, 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(7.6294e-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(0.0035, 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(0.0032, device='cuda:0', grad_fn=<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(0.0006, 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(0.0011, 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(1.6835e-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(2.9582e-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(0.0001, 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 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
				 {'precision': {'support': 0.8604651162790697, 'agreement': 0.6097560975609756, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.4897959183673469, 'partial': 0.13333333333333333}, 'recall': {'support': 0.9380281690140845, 'agreement': 0.47619047619047616, 'direct_attack': 0.1076923076923077, 'undercutter_attack': 0.48, 'partial': 0.6666666666666666}, 'f1': {'support': 0.8975741239892183, 'agreement': 0.5347593582887701, 'direct_attack': 0.1627906976744186, 'undercutter_attack': 0.48484848484848486, 'partial': 0.2222222222222222}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7191780821917808, 'recall': 0.7191780821917808, 'f1': 0.7191780821917809, 'support': None}, 'macro_avg': {'precision': 0.4853367597748117, 'recall': 0.533715523912707, 'f1': 0.4604389774046229, 'support': None}, 'weighted_avg': {'precision': 0.7137773443630243, 'recall': 0.7191780821917808, 'f1': 0.7048156955954705, 'support': None}}
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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2552e-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(9.7363e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9341e-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(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-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(0.0019, 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(4.7716e-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(0.0009, 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(0.0013, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3731e-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(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-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(0.0007, 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(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3862e-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(1.7380e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1531e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6900e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-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(1.5927e-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(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0005, 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(0.0003, 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(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(4.9564e-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(0.0005, 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(0.0030, 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(5.8525e-05, device='cuda:0', grad_fn=<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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, 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(1.6290e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, 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(0.0002, 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(2.0044e-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(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6256e-05, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6163e-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(2.9642e-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(1.9439e-05, 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(0.0004, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9382e-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(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2210e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-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(7.1151e-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(2.8007e-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(0.0006, 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(0.0004, 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(0.0004, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5255e-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(0.0008, 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(0.0006, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-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(3.3397e-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(2.5616e-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(0.0052, 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(3.7151e-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(0.0010, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-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(0.0006, 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(0.0011, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9826e-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(2.4919e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8385e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1751e-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(0.0031, 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(5.5831e-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(0.0005, 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(0.0018, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0009, 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(0.0006, 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(0.0009, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-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(0.0013, 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(4.1057e-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(4.0663e-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(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(0.0008, 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(0.0003, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-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(0.0006, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, 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(0.0044, 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(0.0025, 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(0.0002, 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(0.0006, 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(7.3994e-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(0.0006, 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(0.0009, 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(0.0008, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-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(3.4577e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-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(0.0013, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-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(0.0015, 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(5.1955e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8785e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9957e-05, device='cuda:0', grad_fn=<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(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-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(2.8280e-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(3.8998e-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(8.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-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(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8482, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-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(1.9832e-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(9.5243e-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(0.0004, 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(9.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-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(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-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(0.0015, 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(0.0008, 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(6.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2625e-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(0.0004, 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(0.0051, 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(0.0010, 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(0.0005, 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(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-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(7.4024e-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(1.5836e-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(4.5356e-05, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-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(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(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, 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(2.7281e-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(6.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-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(4.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5508e-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(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(0.0019, 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(0.0001, 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(0.0021, 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(0.0029, 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(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-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(0.0012, 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(1.4352e-05, 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(9.8784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-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(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-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(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8454e-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(0.0010, 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(0.0004, 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(0.0003, 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(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-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(0.0052, 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(0.0009, 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(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-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(2.0287e-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(0.0003, 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(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(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(2.0256e-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(4.0330e-05, 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(0.0003, 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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981e-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(0.0003, 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(7.2268e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3642e-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(3.5819e-05, 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(0.0007, 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(0.0053, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2026e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9655e-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(0.0012, 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(4.4266e-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(2.6887e-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(0.0033, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5773e-05, 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(0.0001, 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(0.0009, 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(0.0028, 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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-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(0.0002, 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(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6001e-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(2.4919e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-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(0.0003, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.1044e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0024, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-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(0.0004, 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(0.0003, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7121e-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(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(0.0052, 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(0.0015, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-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(2.4889e-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(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-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(1.7834e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<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(1.8258e-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(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, device='cuda:0', grad_fn=<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(6.2068e-05, device='cuda:0', grad_fn=<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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, 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(0.0008, 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(0.0664, 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(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-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(0.0008, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(0.0004, 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(2.4616e-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(9.5067e-05, 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(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-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(4.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6434e-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(2.7008e-05, device='cuda:0', grad_fn=<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(1.7531e-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(0.0041, 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(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-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(0.0022, 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(0.0002, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9517e-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(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(3.4093e-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(0.0018, 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(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, 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(0.0004, 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(1.5018e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, 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(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-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(0.0016, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7002e-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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-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(0.0008, 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(0.0004, 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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2975e-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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7718e-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(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0007, 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(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1229e-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(5.6256e-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(0.0011, 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(0.0008, 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(5.0079e-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(0.0004, 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(9.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.1225e-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(6.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-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(9.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2954e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2895e-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(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, 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(0.0004, 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(0.0002, 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(4.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-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(6.9697e-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(4.6234e-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(1.9318e-05, device='cuda:0', grad_fn=<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(5.3378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-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(9.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5809e-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(4.1117e-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(0.0006, 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(8.7818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2029e-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(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8736842105263158, 'agreement': 0.6097560975609756, 'direct_attack': 0.29411764705882354, 'undercutter_attack': 0.4603174603174603, 'partial': 0.11904761904761904}, 'recall': {'support': 0.9352112676056338, 'agreement': 0.47619047619047616, 'direct_attack': 0.07692307692307693, 'undercutter_attack': 0.58, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9034013605442176, 'agreement': 0.5347593582887701, 'direct_attack': 0.12195121951219513, 'undercutter_attack': 0.5132743362831858, 'partial': 0.196078431372549}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7208904109589042, 'recall': 0.7208904109589042, 'f1': 0.7208904109589042, 'support': None}, 'macro_avg': {'precision': 0.4713846069022389, 'recall': 0.5247760752549485, 'f1': 0.4538929412001835, 'support': None}, 'weighted_avg': {'precision': 0.7147041671692974, 'recall': 0.7208904109589042, 'f1': 0.705843266401238, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-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(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1047e-05, 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(0.0004, 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(0.0005, 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(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-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(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6436e-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(0.0007, 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(0.0019, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-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(3.8211e-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(0.0002, 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(1.5291e-05, 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(0.0005, 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(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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, 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(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-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(3.1550e-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(5.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-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(2.2467e-05, 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(4.2813e-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(2.4707e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3711e-05, 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(0.0045, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3524e-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(0.0005, 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(2.0589e-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(0.0004, 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(2.4313e-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(7.3996e-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(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-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(9.4916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-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(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1460e-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(0.0003, 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(1.5563e-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(0.0017, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-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(0.0013, 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(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8674e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8151e-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(0.0036, device='cuda:0', grad_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(1.4958e-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(0.0005, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3479e-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(0.0014, 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(6.3612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3452e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-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(2.1255e-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(0.0012, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0616e-05, 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(0.0001, 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(1.7440e-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(0.0042, 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(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3084e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-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(0.0047, 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(0.0006, 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(5.3530e-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(0.0037, 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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-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(2.5706e-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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(2.0983e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-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(0.0021, 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(0.0003, 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(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5951e-05, 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(9.8121e-05, 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(0.0002, 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(0.0004, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3983e-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(0.0008, 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(0.0011, 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(1.6986e-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(4.6718e-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(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(0.0003, 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(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9785e-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(0.0010, 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(2.2467e-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(2.9854e-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(0.0007, 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(0.0006, 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(2.4102e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8611e-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(0.0026, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0010, 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(2.3496e-05, 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(0.0005, 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(9.5396e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5606e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-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(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-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(5.1652e-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(0.0004, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5669e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6385e-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(1.8016e-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(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(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(0.0017, 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(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, 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(0.0006, 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(0.0008, 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(0.0013, 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(4.1692e-05, device='cuda:0', grad_fn=<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(0.0009, 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(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(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5360e-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(8.1197e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0077e-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(0.0009, 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(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, 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(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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7474e-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(2.3133e-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(7.4057e-05, 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(0.0008, 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(0.0002, 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(0.0011, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3753e-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(0.0003, 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(0.0007, 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(0.0001, 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(0.0009, 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(1.4715e-05, 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(2.5222e-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(0.0049, 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(0.0003, 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(3.1095e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0008, 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(0.0036, 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(0.0015, 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(0.0008, 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(0.0003, 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(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(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5532e-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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-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(0.0004, 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(6.0144e-05, device='cuda:0', grad_fn=<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(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(0.0001, 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(6.0039e-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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1795e-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(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-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(4.7294e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8364e-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(0.0011, 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(0.0014, 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(7.8752e-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(5.8919e-05, 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(3.0278e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-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(2.3829e-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(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(4.9655e-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(0.0007, 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(0.0025, 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(0.0001, 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(7.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0004, 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(0.0049, 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(8.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-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(0.0022, 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(9.0534e-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(4.8384e-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(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(2.9249e-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(2.6009e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, 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(0.0003, 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(0.0002, 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(1.1264e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-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(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-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(0.0011, 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(6.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0006, 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(0.0003, 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(0.0004, 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(5.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-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(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0527e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(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(5.3833e-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(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0001, device='cuda:0', grad_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(8.1803e-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(0.0003, 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(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-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(3.8907e-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(1.5321e-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(2.0892e-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(0.0016, 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(4.5113e-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(7.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-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(0.0009, 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(0.0007, 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(0.0003, 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(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(0.0019, 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(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0930e-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(0.0010, 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(0.0008, 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(0.0002, 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(8.9522e-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(8.8948e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0837e-05, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-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(0.0045, 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(2.6978e-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(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(2.9188e-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(0.0006, 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(2.3920e-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(3.2125e-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(2.0226e-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(2.3678e-05, 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(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(2.8522e-05, device='cuda:0', grad_fn=<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(0.0001, 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(5.4226e-05, 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(6.6244e-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(8.6223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9070e-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(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-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(0.0006, 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(0.0007, 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(0.0013, 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(6.5034e-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(6.8938e-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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-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(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2259e-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(3.6122e-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(0.0046, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-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(3.4154e-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(5.9917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-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(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-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(0.0004, 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(2.5494e-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(6.1794e-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(9.1978e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5349e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-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(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, 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(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5762e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-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(0.0007, 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(0.0014, 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(5.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-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(8.3797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<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(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4498e-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(6.1218e-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(2.3163e-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(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(8.0803e-05, device='cuda:0', grad_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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4295e-05, 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(3.4275e-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(0.0001, 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(0.0003, 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(9.3947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-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(0.0003, 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(0.0006, 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(0.0016, 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
				 {'precision': {'support': 0.8779840848806366, 'agreement': 0.6071428571428571, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45161290322580644, 'partial': 0.11904761904761904}, 'recall': {'support': 0.9323943661971831, 'agreement': 0.4857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9043715846994536, 'agreement': 0.5396825396825397, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5, 'partial': 0.196078431372549}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7208904109589042, 'recall': 0.7208904109589042, 'f1': 0.7208904109589042, 'support': None}, 'macro_avg': {'precision': 0.47431538759622593, 'recall': 0.5251943799549433, 'f1': 0.4565979397223369, 'support': None}, 'weighted_avg': {'precision': 0.7185149651623605, 'recall': 0.7208904109589042, 'f1': 0.7085085606216437, 'support': None}}
Loss: tensor(4.7929e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(7.9939e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3105e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-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(6.4579e-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(1.6593e-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(7.1694e-05, 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(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(0.0003, 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(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, 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(0.0004, 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(1.4473e-05, 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(7.3998e-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(2.7099e-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(2.2315e-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(0.0001, 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(0.0007, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3239e-05, 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(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-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(0.0009, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-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(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(7.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8774e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3641e-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(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(3.6848e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5386e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9241e-05, 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(0.0001, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, 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(5.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-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(6.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-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(1.9832e-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(0.0004, 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(1.0476e-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(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(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-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(0.0013, 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(1.8985e-05, 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(0.0009, 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(9.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0007, 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(0.0005, 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(6.1523e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1403e-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(0.0005, 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(0.0005, 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(0.0004, 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(8.4996e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3469e-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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2966e-05, 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(7.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(1.9499e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-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(4.3175e-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(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, 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(0.0012, 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(4.1389e-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(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4610e-05, 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(8.7972e-05, device='cuda:0', grad_fn=<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(2.3435e-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(8.7707e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, 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(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(7.2150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-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(0.0001, 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(0.0007, 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(1.1657e-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(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-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(0.0009, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-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(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-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(6.7855e-05, device='cuda:0', grad_fn=<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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-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(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-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(1.9378e-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(0.0023, 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(0.0007, 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(0.0018, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-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(3.0914e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5173e-05, device='cuda:0', grad_fn=<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(6.0523e-05, 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(0.0025, 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(0.0001, 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(3.6455e-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(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(6.6638e-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(1.7955e-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(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-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(0.0003, 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(0.0019, 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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-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(0.0005, 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(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(0.0003, 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(4.0905e-05, device='cuda:0', grad_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(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(9.9837e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5841e-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(1.7955e-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(4.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7235e-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(2.6736e-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(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1309e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7154e-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(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-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(5.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, 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(0.0018, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0299e-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(2.9006e-05, device='cuda:0', grad_fn=<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(7.8023e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0035, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-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(0.0022, 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(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-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(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(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2378e-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(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0020, 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(7.6246e-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(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4337e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6899e-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(2.2134e-05, 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(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(4.5265e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-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(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(0.0004, 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(2.7795e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, 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(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(2.9612e-05, device='cuda:0', grad_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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-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(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(0.0012, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0026, 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(2.1286e-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(2.0650e-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(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2793e-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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0034, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-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(1.7562e-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(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(9.3814e-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(0.0003, 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(2.0953e-05, device='cuda:0', grad_fn=<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(8.8194e-05, 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(0.0029, 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(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(4.8353e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, 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(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1167e-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(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6484e-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(0.0015, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2351e-05, 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(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, 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(1.6411e-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(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(4.5659e-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(0.0011, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4356e-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(1.9530e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9059e-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(5.8826e-05, 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(0.0002, 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(0.0018, 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(0.0002, 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(1.7864e-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(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-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(0.0009, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-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(0.0003, 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(1.0900e-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(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(0.0012, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-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(6.4065e-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(6.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-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(0.0001, 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(7.9634e-06, 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(7.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8163e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-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(1.8167e-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(9.8489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4819e-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(0.0010, 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(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7424e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0031, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9458e-05, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-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(0.0010, 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(0.0011, 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(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0300e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-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(0.0002, 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(2.1286e-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(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4701e-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(0.0013, 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(1.5260e-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(2.6100e-05, 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(2.1952e-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(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(6.2308e-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(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-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(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3205e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8058e-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(2.2012e-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(0.0001, 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(0.0001, 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(2.7372e-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(0.0006, 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(0.0006, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-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(2.7523e-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(0.0011, 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(1.7985e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-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(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2439e-05, 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(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8192e-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(2.1074e-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(2.3133e-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(0.0009, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5508e-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(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1310e-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(0.0007, 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(0.0008, 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(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-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(1.9802e-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(1.3111e-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(0.0001, 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(0.0001, 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(0.0024, 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(0.0021, 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(4.6808e-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(1.7410e-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(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5135e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3143e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, 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(0.0011, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-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(4.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3076e-05, 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(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(3.8332e-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(8.2171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(6.2036e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-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(2.6221e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<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(0.0004, 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(7.4931e-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(4.2661e-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(0.0030, device='cuda:0', grad_fn=<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(1.7895e-05, 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(0.0010, 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(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8573e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-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(0.0003, 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(0.0023, 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(0.0010, 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
				 {'precision': {'support': 0.8823529411764706, 'agreement': 0.6144578313253012, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45161290322580644, 'partial': 0.10869565217391304}, 'recall': {'support': 0.9295774647887324, 'agreement': 0.4857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9053497942386831, 'agreement': 0.5425531914893618, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5, 'partial': 0.1818181818181818}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7191780821917808, 'recall': 0.7191780821917808, 'f1': 0.7191780821917809, 'support': None}, 'macro_avg': {'precision': 0.4745817603171404, 'recall': 0.5246309996732531, 'f1': 0.4545156620806739, 'support': None}, 'weighted_avg': {'precision': 0.7223263497039947, 'recall': 0.7191780821917808, 'f1': 0.7093995547657421, 'support': None}}
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(0.0002, 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(3.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<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(0.0001, 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(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3056e-05, 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(0.0022, 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(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, 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(9.5364e-05, device='cuda:0', grad_fn=<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(4.7203e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5802e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2127e-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(0.0001, 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(7.5766e-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(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3955e-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(8.5949e-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(4.4295e-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(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(0.0005, 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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1773e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9211e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, 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(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-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(1.9348e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6375e-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(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(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(1.4958e-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(1.3141e-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(2.9219e-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(0.0005, 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(0.0003, 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(2.1013e-05, 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(7.4477e-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(5.9675e-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(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-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(0.0006, 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(1.6532e-05, 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(6.6276e-05, device='cuda:0', grad_fn=<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(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, 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(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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5043e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3321e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-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(0.0001, device='cuda:0', grad_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(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-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(0.0005, 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(8.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-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(0.0014, 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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, 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(8.7836e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5030e-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(1.4897e-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(0.0004, 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(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(0.0017, 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(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-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(0.0007, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, 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(2.2315e-05, 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(1.6744e-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(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(0.0006, 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(1.6986e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5426e-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(1.3141e-05, 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(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-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(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9876e-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(0.0011, 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(0.0006, 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(2.6705e-05, 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(3.8392e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-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(3.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6668e-05, 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(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(0.0009, 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(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(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(6.4881e-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(0.0005, 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(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(0.0003, 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(2.0771e-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(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5649e-05, device='cuda:0', grad_fn=<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(0.0007, 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(6.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-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(0.0004, 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(1.1839e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-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(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4863e-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(7.3389e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-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(2.9733e-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(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, 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(0.0005, 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(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-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(0.0003, 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(9.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-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(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(0.0003, 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(0.0002, 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(6.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1663e-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(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3499e-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(0.0002, 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(5.9886e-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(0.0011, 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(2.8734e-05, device='cuda:0', grad_fn=<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(1.3504e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0004, 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(0.0008, 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(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2369e-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(0.0009, 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(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-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(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0405e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-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(1.9197e-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(0.0005, 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(8.6055e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-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(0.0003, 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(1.1233e-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(4.1450e-05, device='cuda:0', grad_fn=<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(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7374e-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(0.0007, 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(0.0018, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-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(0.0015, 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(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, 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(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(0.0002, 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(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-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(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-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(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(1.5624e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(0.0004, 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(0.0016, 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(0.0003, 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(6.0825e-05, device='cuda:0', grad_fn=<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(2.0226e-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(5.7555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-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(2.7553e-05, device='cuda:0', grad_fn=<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(2.9733e-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(0.0003, 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(0.0004, 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(1.6350e-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(4.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-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(0.0007, 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(1.7380e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1239e-05, device='cuda:0', grad_fn=<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(2.2981e-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(0.0002, 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(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6729e-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(0.0003, 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(9.6849e-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(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(1.7683e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6912e-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(0.0001, 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(0.0002, 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(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-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(0.0003, 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(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-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(7.0364e-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(0.0010, 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(0.0001, 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(1.1082e-05, 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(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(2.2194e-05, device='cuda:0', grad_fn=<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(1.4837e-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(1.3868e-05, device='cuda:0', grad_fn=<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(1.0204e-05, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-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(9.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, 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(0.0009, 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(8.1992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-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(0.0005, 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(6.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-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(3.9452e-05, device='cuda:0', grad_fn=<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(1.3595e-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(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7366e-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(4.9867e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-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(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, 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(5.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2446e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8239e-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(6.2113e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5960e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4482e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-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(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5912e-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(5.1925e-05, device='cuda:0', grad_fn=<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(1.0265e-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(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2257e-05, device='cuda:0', grad_fn=<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(2.8038e-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(0.0003, 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(0.0003, 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(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-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(1.4806e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-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(0.0002, 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(0.0001, 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(0.0017, 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(0.0002, 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(0.0003, 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(3.6273e-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(8.5441e-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(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_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(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9745e-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(0.0003, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4496e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9501e-05, device='cuda:0', grad_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(2.0983e-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(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(2.9642e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-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(2.3859e-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(0.0023, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0998e-05, 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(0.0004, 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(2.0408e-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(2.5615e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-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(2.0014e-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(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-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(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-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(0.0002, 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(0.0002, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-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(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(2.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-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(0.0015, 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(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.1900e-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(5.2803e-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(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, 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(0.0009, 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(1.0295e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0805e-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(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6405e-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(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1486e-05, device='cuda:0', grad_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(0.0004, 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(1.7017e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-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(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2749e-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(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1836e-05, device='cuda:0', grad_fn=<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(2.4919e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1400e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, 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(0.0019, 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(0.0005, 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(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-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(8.1753e-06, 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(3.1671e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0693e-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(0.0018, 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(1.2414e-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(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0676e-05, 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(2.0135e-05, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8820375335120644, 'agreement': 0.6071428571428571, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10638297872340426}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9038461538461539, 'agreement': 0.5396825396825397, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.17857142857142855}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7174657534246576, 'recall': 0.7174657534246576, 'f1': 0.7174657534246577, 'support': None}, 'macro_avg': {'precision': 0.47407384730103186, 'recall': 0.5240676193915631, 'f1': 0.45389235389235383, 'support': None}, 'weighted_avg': {'precision': 0.7214176484022231, 'recall': 0.7174657534246576, 'f1': 0.7083050233735165, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<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(5.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-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(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(0.0006, 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(0.0001, 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(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(7.9743e-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(9.4117e-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(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(3.1883e-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(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-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(0.0009, 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(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5739e-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(3.2125e-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(0.0007, 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(0.0006, 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(0.0001, 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(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-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(3.4759e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2445e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3109e-05, device='cuda:0', grad_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(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0015, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0947e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-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(0.0001, 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(3.5668e-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(2.7250e-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(0.0005, 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(0.0001, 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(0.0003, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1817e-05, device='cuda:0', grad_fn=<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(8.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-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(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9544e-05, device='cuda:0', grad_fn=<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(2.6554e-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(0.0002, 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(9.5272e-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(0.0016, 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(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-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(0.0002, 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(3.0338e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-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(0.0004, 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(8.6677e-05, 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(1.4655e-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(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(0.0004, 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(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(9.9431e-05, 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(1.0779e-05, device='cuda:0', grad_fn=<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(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, 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(8.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-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(0.0007, 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(0.0003, 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(5.0421e-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(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2471e-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(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(9.8532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4093e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(8.4866e-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(8.8917e-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(0.0006, 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(2.3799e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1634e-05, 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(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0253e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5345e-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(0.0002, 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(1.5260e-05, device='cuda:0', grad_fn=<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(2.9249e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9242e-05, 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(0.0001, 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(0.0002, 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(0.0011, 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(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0830e-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(4.4568e-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(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(1.6290e-05, 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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7016e-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(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-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(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0006, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-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(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(0.0001, 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(2.8946e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3718e-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(5.4773e-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(0.0003, 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(2.4041e-05, device='cuda:0', grad_fn=<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(2.3859e-05, device='cuda:0', grad_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(1.0355e-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(1.4019e-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(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(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(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8088e-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(1.8349e-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(1.5260e-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(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1908e-05, 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(0.0005, 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(0.0005, 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(2.3284e-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(2.4919e-05, device='cuda:0', grad_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(2.2012e-05, 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(0.0001, 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(8.9963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(4.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-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(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-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(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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8992e-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(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8191e-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(3.0338e-05, device='cuda:0', grad_fn=<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(3.6074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-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(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-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(0.0020, 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(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(8.7161e-05, 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(0.0011, 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(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2721e-05, device='cuda:0', grad_fn=<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(1.1597e-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(0.0003, 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(4.4750e-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(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1922e-05, device='cuda:0', grad_fn=<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(5.7067e-05, 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(0.0002, 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(0.0002, 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(0.0002, 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(1.0265e-05, 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(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(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-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(1.5896e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5790e-05, 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(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(1.2657e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2596e-05, device='cuda:0', grad_fn=<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(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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8291e-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(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8807e-05, 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(0.0001, 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(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4043e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9896e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7746e-05, device='cuda:0', grad_fn=<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(0.0016, 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(1.3202e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-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(1.0961e-05, 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(9.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4575e-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(0.0016, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-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(0.0001, 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(8.3947e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8151e-05, device='cuda:0', grad_fn=<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(2.5888e-05, 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(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5431e-05, device='cuda:0', grad_fn=<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(0.0001, 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(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, 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(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0010, 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(1.5745e-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(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4779e-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(1.2687e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6435e-05, device='cuda:0', grad_fn=<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(4.0784e-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(0.0005, 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(0.0019, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-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(9.4734e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-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(4.4569e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(2.0468e-05, device='cuda:0', grad_fn=<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(6.7078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5349e-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(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3742e-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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-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(0.0006, 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(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7923e-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(1.7804e-05, device='cuda:0', grad_fn=<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(0.0006, 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(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-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(8.1502e-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(1.5775e-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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6515e-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(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(7.1579e-05, device='cuda:0', grad_fn=<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(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-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(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-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(9.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5871e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4718e-05, 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(1.2323e-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(1.8228e-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(2.2648e-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(5.1864e-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(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4270e-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(1.6714e-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(0.0003, 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(0.0002, 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(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0117e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7785e-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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3852e-05, device='cuda:0', grad_fn=<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(5.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4770e-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(8.2804e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8685e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-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(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9260e-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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6616e-05, 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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7709e-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(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(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3288e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8903e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8132e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-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(1.6350e-05, device='cuda:0', grad_fn=<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(1.9287e-05, 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(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4731e-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(1.4080e-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(0.0001, 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(0.0003, 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(0.0002, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-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(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(3.7726e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6699e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0001, 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(0.0005, 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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1184e-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(0.0006, 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(0.0004, 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(7.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7733e-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(4.6264e-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(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-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(4.7986e-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(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(0.0004, 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(1.7047e-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(5.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4050e-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(2.3799e-05, device='cuda:0', grad_fn=<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(7.9844e-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(4.8553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, 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(8.2016e-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(0.0009, 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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-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(8.2661e-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(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7857e-05, device='cuda:0', grad_fn=<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(6.4429e-05, 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(1.7289e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6609e-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(5.5983e-05, device='cuda:0', grad_fn=<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(2.6191e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4508e-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(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4339e-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(1.4080e-05, 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(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, 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(0.0002, 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
				 {'precision': {'support': 0.8820375335120644, 'agreement': 0.6144578313253012, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10416666666666667}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9038461538461539, 'agreement': 0.5425531914893618, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.1754385964912281}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7174657534246576, 'recall': 0.7174657534246576, 'f1': 0.7174657534246577, 'support': None}, 'macro_avg': {'precision': 0.47509357972617317, 'recall': 0.5240676193915631, 'f1': 0.4538399178376783, 'support': None}, 'weighted_avg': {'precision': 0.7226986851841513, 'recall': 0.7174657534246576, 'f1': 0.7087728708923429, 'support': None}}
Loss: tensor(7.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4040e-05, device='cuda:0', grad_fn=<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(4.6748e-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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7859e-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(3.7726e-05, device='cuda:0', grad_fn=<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(0.0004, 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(1.2414e-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(0.0010, 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(8.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, 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(6.0766e-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(5.6374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, 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(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(7.4741e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(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(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(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5248e-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(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, 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(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6824e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-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(3.3911e-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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7989e-05, 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(1.4746e-05, 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(2.0347e-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(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-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(1.2475e-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(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(0.0006, 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(3.9663e-05, 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(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(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-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(4.9364e-05, device='cuda:0', grad_fn=<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(7.5506e-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(7.5697e-06, 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(2.2800e-05, 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(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9311e-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(0.0004, 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(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6055e-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(1.2233e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, 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(2.1104e-05, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0759e-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(0.0002, device='cuda:0', grad_fn=<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(7.7171e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3452e-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(2.4889e-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(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(0.0001, 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(0.0002, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, 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(1.5018e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4012e-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(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5882e-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(0.0001, 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(0.0007, 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(3.3124e-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(7.4486e-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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-05, 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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2216e-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(7.1682e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4543e-05, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(2.2951e-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(7.5438e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1350e-05, 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(3.4941e-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(4.5257e-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(2.6069e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-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(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0714e-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(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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2802e-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(0.0007, 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(0.0015, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-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(1.5412e-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(0.0002, 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(5.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3660e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, 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(0.0001, 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(1.6411e-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(0.0001, 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(3.4729e-05, device='cuda:0', grad_fn=<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(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(9.2351e-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(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-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(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(0.0009, 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(6.0660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-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(6.8207e-05, device='cuda:0', grad_fn=<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(7.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2768e-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(3.6163e-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(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-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(0.0005, 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(3.6303e-05, device='cuda:0', grad_fn=<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(5.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8243e-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(4.3326e-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(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(2.3890e-05, 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(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1953e-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(8.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, 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(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-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(0.0002, 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(1.2172e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0001, 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(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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3759e-05, 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(7.9634e-06, 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(0.0003, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(8.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3857e-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(0.0003, 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(1.3141e-05, 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(0.0008, 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(2.0680e-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>)
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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-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(5.5024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3810e-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(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(8.9323e-06, 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(1.3474e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1882e-05, device='cuda:0', grad_fn=<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(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8352e-05, device='cuda:0', grad_fn=<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(7.7025e-05, device='cuda:0', grad_fn=<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(1.5896e-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(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0756e-05, 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(9.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-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(1.5139e-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(0.0001, device='cuda:0', grad_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(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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9484e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3610e-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(0.0004, 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(0.0013, 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(6.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0378e-05, 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(2.5646e-05, device='cuda:0', grad_fn=<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(1.3837e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7953e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6055e-05, device='cuda:0', grad_fn=<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(0.0008, 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(4.0238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5567e-05, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5145e-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(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-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(7.3704e-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(0.0001, 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(9.6892e-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(0.0004, 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(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-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(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<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(5.0592e-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(1.5018e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-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(0.0009, 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(8.0239e-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(0.0005, 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(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0422e-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(4.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0883e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5322e-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(0.0004, 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(1.1869e-05, device='cuda:0', grad_fn=<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(1.6169e-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(7.0807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9729e-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(0.0002, 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(0.0004, 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(2.4465e-05, 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(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(2.7583e-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(1.8409e-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(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-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(0.0006, 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(9.3879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6966e-05, device='cuda:0', grad_fn=<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(9.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-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(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, 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(5.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3389e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-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(8.2227e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3039e-05, device='cuda:0', grad_fn=<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(1.0931e-05, device='cuda:0', grad_fn=<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(4.5810e-05, 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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-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(0.0003, 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(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-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(0.0011, 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(0.0005, 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(5.9249e-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(4.3630e-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(1.7259e-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(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(0.0004, 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(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3162e-05, 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(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7337e-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(8.1450e-06, 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(1.5805e-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(2.0892e-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(7.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1180e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9720e-05, 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(0.0001, 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(7.5437e-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(9.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_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(1.2687e-05, 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(0.0002, 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(0.0012, 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(3.3941e-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(0.0001, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, 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(1.4443e-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(7.4486e-06, 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(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-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(7.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0001, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, 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(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(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1624e-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(5.6492e-05, device='cuda:0', grad_fn=<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(1.0386e-05, 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(4.8959e-05, device='cuda:0', grad_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(0.0004, 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(1.0931e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2418e-05, device='cuda:0', grad_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(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(9.3864e-06, 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(1.1052e-05, device='cuda:0', grad_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(7.7838e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2450e-05, device='cuda:0', grad_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(5.3271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, 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(1.4110e-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(8.7506e-06, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1474e-05, 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(9.3911e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0005, 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(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-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(7.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0823e-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(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<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(6.8934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1428e-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(0.0002, 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(1.5109e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5094e-05, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-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(0.0003, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-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(0.0004, 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(8.4781e-06, 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(0.0021, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9193e-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(2.5827e-05, device='cuda:0', grad_fn=<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(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(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(9.9178e-05, 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(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-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(7.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3349e-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(4.0239e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3015e-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(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6745e-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(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-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(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6980e-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
				 {'precision': {'support': 0.8820375335120644, 'agreement': 0.6219512195121951, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10204081632653061}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9038461538461539, 'agreement': 0.5454545454545454, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.1724137931034483}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7174657534246576, 'recall': 0.7174657534246576, 'f1': 0.7174657534246577, 'support': None}, 'macro_avg': {'precision': 0.4761670872955247, 'recall': 0.5240676193915631, 'f1': 0.45381522795315893, 'support': None}, 'weighted_avg': {'precision': 0.7240131939282654, 'recall': 0.7174657534246576, 'f1': 0.7092479033167508, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-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(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, 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(0.0002, 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(2.2618e-05, 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(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(4.5264e-05, 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(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3508e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-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(9.8709e-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(0.0005, 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(0.0002, 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(0.0005, 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(9.7246e-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(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-05, device='cuda:0', grad_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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.8009e-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(3.5637e-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(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3910e-05, device='cuda:0', grad_fn=<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(1.0779e-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(0.0003, 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(5.1999e-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(8.0263e-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(1.9136e-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(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4013e-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(3.4535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210e-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(2.2164e-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(1.5927e-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(4.0885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-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(6.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-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(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-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(6.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5610e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-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(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-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(0.0004, 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(8.6310e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-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(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(9.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6551e-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(0.0002, 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(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, 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(3.1883e-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(0.0003, 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(7.5447e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6151e-05, device='cuda:0', grad_fn=<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(9.1293e-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(3.4093e-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(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(5.2838e-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(9.6695e-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(9.9349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9786e-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(0.0003, 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(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3892e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9899e-05, device='cuda:0', grad_fn=<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(1.7380e-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(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, 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(8.5244e-05, device='cuda:0', grad_fn=<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(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6803e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0001, 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(1.0355e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3761e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2929e-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(0.0001, 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(7.3951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2756e-05, device='cuda:0', grad_fn=<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(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6804e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7008e-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(6.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8608e-05, device='cuda:0', grad_fn=<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(4.6632e-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(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(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8694e-05, device='cuda:0', grad_fn=<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(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, 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(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-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(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(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-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(8.8825e-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(1.0870e-05, device='cuda:0', grad_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(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(0.0001, 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(0.0002, 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(0.0007, 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(6.6075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7500e-05, 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(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9401e-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(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9230e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3444e-05, device='cuda:0', grad_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(9.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6054e-05, 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(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(9.5076e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-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(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<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(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, 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(1.0688e-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(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7226e-05, device='cuda:0', grad_fn=<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(9.2956e-06, 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(0.0001, 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(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9369e-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(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4379e-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(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0703e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3701e-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(0.0002, 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(1.2445e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-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(8.3423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1367e-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(8.6344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6243e-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(0.0002, 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(7.6323e-05, device='cuda:0', grad_fn=<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(1.2414e-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(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<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(0.0001, 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(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-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(0.0012, 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(1.2899e-05, 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(0.0002, 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(0.0012, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3587e-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(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7710e-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(2.4767e-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(1.0325e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-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(8.8427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-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(6.5403e-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(2.9612e-05, device='cuda:0', grad_fn=<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(4.7354e-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(8.4821e-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(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1164e-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(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-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(7.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, 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(1.3353e-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(0.0003, 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(1.9318e-05, 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(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(2.3042e-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(2.8946e-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(3.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2796e-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(1.1355e-05, 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(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4830e-05, device='cuda:0', grad_fn=<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(8.4920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, 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(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(1.0961e-05, 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(1.4443e-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(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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, 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(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(8.5689e-06, 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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8454e-05, 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(0.0001, 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(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3900e-05, 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(0.0006, 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(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9202e-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(7.1874e-05, device='cuda:0', grad_fn=<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(7.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, 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(8.3873e-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(2.8552e-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(3.3760e-05, 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(4.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1669e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.3716e-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(1.5230e-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(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5007e-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(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4137e-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(6.6008e-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(6.7825e-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(6.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, 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(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(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5985e-05, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-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(7.9035e-05, device='cuda:0', grad_fn=<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(7.1749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, 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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6767e-05, 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(8.0342e-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(9.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6048e-05, 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(3.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.5593e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2792e-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(5.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(8.9323e-06, 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(0.0007, 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(0.0004, 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(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-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(8.7012e-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(7.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8419e-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(7.9967e-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(2.5948e-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(1.6048e-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(7.3578e-06, 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(3.5849e-05, 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(5.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-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(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8930e-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(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(3.6586e-05, 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(0.0001, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4883e-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(2.7674e-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(7.1107e-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(0.0002, 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(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8069e-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(0.0001, 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(7.4599e-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(7.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9627e-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(4.2055e-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(0.0001, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-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(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-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(6.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4558e-05, 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(7.4789e-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(1.0204e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-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(7.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9251e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8908e-05, device='cuda:0', grad_fn=<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(0.0003, 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(7.7495e-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(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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-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(7.9351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9203e-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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7518e-05, 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(0.0004, 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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4157e-05, 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(0.0002, 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(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(9.0498e-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(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3136e-05, 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(6.3586e-06, 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(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9381e-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(3.9243e-05, device='cuda:0', grad_fn=<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(8.8768e-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(0.0003, 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(7.7499e-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(0.0005, 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(0.0005, 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(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(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<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(5.5589e-05, device='cuda:0', grad_fn=<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(1.3262e-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(0.0005, 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(0.0003, 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
				 {'precision': {'support': 0.8820375335120644, 'agreement': 0.6219512195121951, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10204081632653061}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4857142857142857, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9038461538461539, 'agreement': 0.5454545454545454, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.1724137931034483}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.7174657534246576, 'recall': 0.7174657534246576, 'f1': 0.7174657534246577, 'support': None}, 'macro_avg': {'precision': 0.4761670872955247, 'recall': 0.5240676193915631, 'f1': 0.45381522795315893, 'support': None}, 'weighted_avg': {'precision': 0.7240131939282654, 'recall': 0.7174657534246576, 'f1': 0.7092479033167508, 'support': None}}
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>)
Loss: tensor(0.0003, 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(6.2126e-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(0.0010, 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(0.0004, 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(0.0013, 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(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5880e-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(5.9250e-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(6.5705e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1014e-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(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(0.0006, 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(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7347e-05, device='cuda:0', grad_fn=<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(4.1764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-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(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(9.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3378e-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(1.1536e-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(0.0002, device='cuda:0', grad_fn=<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(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, 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(0.0002, 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(1.4080e-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(1.6441e-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(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3223e-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(0.0007, 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(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(0.0003, 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(7.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-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(0.0003, 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(0.0001, 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(0.0004, 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(0.0002, 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(4.2776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4614e-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(9.8696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-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(6.8320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4628e-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(1.3050e-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(0.0003, 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(5.7999e-05, device='cuda:0', grad_fn=<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(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-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(9.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7058e-05, 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(0.0004, 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(7.4486e-06, 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(0.0003, 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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0340e-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(8.7203e-06, 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(8.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9929e-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(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7626e-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(0.0002, 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(8.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6014e-05, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0006, 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(1.7501e-05, device='cuda:0', grad_fn=<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(5.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-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(3.7847e-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(8.1015e-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(0.0003, 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(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7258e-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(7.4788e-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(4.1599e-05, device='cuda:0', grad_fn=<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(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2152e-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(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(1.8561e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7919e-05, device='cuda:0', grad_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(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0003, 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(0.0001, 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(7.5586e-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(3.0520e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-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(8.0239e-06, 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(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-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(3.9669e-05, 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(6.2558e-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(2.9400e-05, 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(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1396e-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(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(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9442e-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(5.5041e-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(0.0007, 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(1.1112e-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(1.1203e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-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(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8715e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-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(7.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4822e-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(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0934e-05, device='cuda:0', grad_fn=<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(0.0003, 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(3.1428e-05, device='cuda:0', grad_fn=<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(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4554e-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(4.2298e-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(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(5.7530e-06, 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(0.0022, 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(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, 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(1.3262e-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(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8151e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9166e-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(7.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1682e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0787e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9461e-05, device='cuda:0', grad_fn=<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(0.0006, 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(5.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, 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(9.7498e-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(7.8120e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-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(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(5.7785e-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(8.1540e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1599e-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(9.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, 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(9.5532e-05, 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(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, 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(8.3267e-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(5.7979e-05, 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(0.0003, 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(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-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(1.3050e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-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(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7534e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4803e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2609e-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(5.5618e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-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(7.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, 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(0.0003, 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(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6920e-05, device='cuda:0', grad_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(0.0005, 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(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3066e-05, device='cuda:0', grad_fn=<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(0.0002, 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(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1886e-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(1.9075e-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(7.3817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-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(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(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-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(0.0009, 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(0.0010, 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(0.0001, 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(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-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(8.6598e-06, 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(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(8.3195e-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(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4443e-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(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, 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(6.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, 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(0.0002, 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(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5345e-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(8.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6981e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9777e-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(6.3334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, 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(0.0010, 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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8349e-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(0.0002, 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(1.5079e-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(7.1458e-06, 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(2.4283e-05, device='cuda:0', grad_fn=<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(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3829e-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(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(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(7.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, 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(8.0202e-05, 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(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1490e-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(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9933e-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(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7005e-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(7.4485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2144e-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(5.4545e-05, device='cuda:0', grad_fn=<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(7.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0548e-05, device='cuda:0', grad_fn=<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(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8656e-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(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9989e-05, 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(0.0006, 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(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0792e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7329e-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(0.0001, 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(9.1139e-06, 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(0.0001, 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(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-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(4.4378e-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(2.4313e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8668e-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(1.4019e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8723e-05, device='cuda:0', grad_fn=<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(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8836e-05, device='cuda:0', grad_fn=<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(5.3594e-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(9.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6657e-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(6.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8606e-05, 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(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2449e-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(1.0537e-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(8.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0683e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, 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(0.0002, 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(3.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9995e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-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(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(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9023e-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(0.0002, 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(2.8431e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2530e-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(3.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-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(0.0002, 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(8.6295e-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(6.6343e-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(5.5705e-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(1.2233e-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(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-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(6.8128e-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(3.0672e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-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(7.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0093e-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(8.5101e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1309e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5424e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8649e-05, device='cuda:0', grad_fn=<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(0.0007, 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(4.5627e-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(2.0317e-05, device='cuda:0', grad_fn=<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(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4403e-05, device='cuda:0', grad_fn=<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(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5351e-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(3.1095e-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(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(1.0022e-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(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(8.2219e-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(7.6370e-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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, 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(3.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4286e-05, device='cuda:0', grad_fn=<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(9.2351e-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(5.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8338e-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(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-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(9.6892e-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(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7208e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0533e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8812e-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(1.7017e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2464e-05, 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(1.6744e-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(0.0002, 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(0.0002, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-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(6.9816e-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
				 {'precision': {'support': 0.8773333333333333, 'agreement': 0.6125, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10204081632653061}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4666666666666667, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9013698630136987, 'agreement': 0.5297297297297299, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.1724137931034483}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.714041095890411, 'recall': 0.714041095890411, 'f1': 0.7140410958904111, 'support': None}, 'macro_avg': {'precision': 0.4733360033573395, 'recall': 0.5202580955820393, 'f1': 0.4501750066417049, 'support': None}, 'weighted_avg': {'precision': 0.7194543427086935, 'recall': 0.714041095890411, 'f1': 0.7049153880828001, 'support': None}}
Loss: tensor(5.2287e-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(1.0204e-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(0.0004, 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(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7156e-05, 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(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4577e-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(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-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(1.0931e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5481e-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(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, 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(2.6675e-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(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, 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(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(0.0002, 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(6.8808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9031e-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(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0340e-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(0.0002, 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(5.8118e-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(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-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(0.0005, 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(4.2035e-05, 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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6793e-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(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1800e-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(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6438e-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(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0996e-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(1.1536e-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(7.4876e-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(3.8681e-05, device='cuda:0', grad_fn=<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(7.7817e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8892e-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(9.5284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9215e-05, device='cuda:0', grad_fn=<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(4.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2794e-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(5.1474e-06, 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(8.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, 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(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(1.6835e-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(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, 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(6.9996e-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(5.6927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3451e-05, 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(3.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-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(8.0482e-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(1.6986e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-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(0.0002, 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(3.6333e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5707e-05, device='cuda:0', grad_fn=<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(7.0356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, 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(4.2844e-05, device='cuda:0', grad_fn=<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(2.1407e-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(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(8.6295e-06, 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(0.0001, 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(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4923e-05, 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(1.3777e-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(2.9370e-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(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(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(8.7554e-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(2.6221e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6794e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1935e-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(4.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7845e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9070e-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(0.0002, 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(9.4773e-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(9.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-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(0.0003, 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(0.0001, 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(7.6612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-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(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(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(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0171e-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(6.3195e-05, device='cuda:0', grad_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(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2357e-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(0.0004, 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(3.2237e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5311e-05, device='cuda:0', grad_fn=<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(1.9197e-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(0.0002, 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(0.0001, 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(1.1203e-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(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(9.6589e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-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(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0020e-05, device='cuda:0', grad_fn=<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(7.7867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3640e-05, device='cuda:0', grad_fn=<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(7.1458e-06, 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(2.6554e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1260e-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(8.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, 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(0.0004, 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(8.5689e-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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0198e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-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(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6155e-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(7.4991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8543e-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(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-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(0.0001, 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(5.3711e-05, 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(5.7943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, 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(0.0002, 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(9.6570e-05, device='cuda:0', grad_fn=<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(6.4780e-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(7.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(5.2080e-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(7.3605e-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(0.0004, 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(9.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-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(7.0197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, 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(9.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3326e-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(6.9339e-06, 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(4.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-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(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0746e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(2.5343e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4467e-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(4.9391e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9566e-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(1.2202e-05, device='cuda:0', grad_fn=<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(4.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5241e-05, 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(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5826e-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(6.1656e-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(0.0003, 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(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-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(0.0002, 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(0.0001, 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(0.0002, 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(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7385e-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(0.0002, 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(0.0001, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7574e-05, device='cuda:0', grad_fn=<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(9.8344e-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(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-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(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0392e-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(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2291e-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(1.1021e-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(9.9920e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, 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(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(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, 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(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-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(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(7.1761e-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(0.0015, 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(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0518e-05, 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(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5993e-05, device='cuda:0', grad_fn=<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(0.0001, 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(8.4478e-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(9.9012e-06, 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(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8125e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5670e-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(0.0002, 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(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6528e-05, 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(6.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1937e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2962e-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(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9534e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4225e-05, 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(0.0003, 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(2.4919e-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(7.0600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-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(9.2048e-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(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2450e-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(3.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, 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(4.1086e-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(0.0003, 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(0.0003, 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(8.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1571e-05, device='cuda:0', grad_fn=<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(7.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6913e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-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(5.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0745e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, 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(0.0003, 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(1.4473e-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(4.3326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, 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(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1619e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2207e-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(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(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, 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(5.1192e-05, 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(8.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, 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(0.0001, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8694e-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(8.3279e-05, device='cuda:0', grad_fn=<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(8.8112e-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(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8476e-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(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7314e-05, device='cuda:0', grad_fn=<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(4.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-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(9.2955e-05, device='cuda:0', grad_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(6.8272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, 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(7.2669e-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(0.0014, 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(7.9859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9648e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(6.2529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2702e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-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(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8603e-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(7.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7875e-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(5.2854e-05, 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(2.4919e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7800e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(4.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3038e-05, 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(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-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(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(1.0174e-05, device='cuda:0', grad_fn=<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(8.1753e-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(4.5324e-05, 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(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3022e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9015e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2326e-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(0.0004, 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(5.1077e-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(0.0003, 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(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(7.2298e-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(6.2708e-05, 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(0.0003, 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(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(9.3159e-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(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, 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(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2813e-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(1.0113e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, 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(0.0001, 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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-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(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-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(9.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, 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(0.0002, 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(8.2918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5309e-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(7.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1113e-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(0.0017, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8773333333333333, 'agreement': 0.6125, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10204081632653061}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4666666666666667, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9013698630136987, 'agreement': 0.5297297297297299, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.1724137931034483}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.714041095890411, 'recall': 0.714041095890411, 'f1': 0.7140410958904111, 'support': None}, 'macro_avg': {'precision': 0.4733360033573395, 'recall': 0.5202580955820393, 'f1': 0.4501750066417049, 'support': None}, 'weighted_avg': {'precision': 0.7194543427086935, 'recall': 0.714041095890411, 'f1': 0.7049153880828001, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5296e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6536e-05, 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(0.0001, 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(1.2445e-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(9.9971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4682e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3447e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9462e-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(7.3009e-05, device='cuda:0', grad_fn=<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(9.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0110e-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(0.0001, 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(8.4180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8366e-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(8.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2562e-05, 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(0.0002, device='cuda:0', grad_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(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1413e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1720e-05, device='cuda:0', grad_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(0.0003, 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(5.7676e-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(0.0002, 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(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-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(0.0001, 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(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7027e-05, 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(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2686e-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(6.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4806e-05, 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(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(7.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1592e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3741e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-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(7.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5638e-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(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(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7134e-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(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6548e-05, 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(4.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8703e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5683e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2771e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1291e-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(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(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4723e-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(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3657e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1415e-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(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6097e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3626e-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(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5650e-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(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, 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(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1792e-05, device='cuda:0', grad_fn=<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(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9672e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-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(0.0007, 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(3.5664e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4064e-05, device='cuda:0', grad_fn=<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(0.0001, 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(6.4306e-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(9.5134e-05, 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(9.1745e-06, 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(9.1500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(7.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0017e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6565e-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(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-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(2.4556e-05, device='cuda:0', grad_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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7959e-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(0.0002, 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(0.0001, 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(0.0004, 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(1.2354e-05, 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(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(1.8773e-05, device='cuda:0', grad_fn=<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(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(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, 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(9.3797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9133e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7724e-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(8.0845e-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(8.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8666e-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(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6651e-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(8.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, 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(6.9088e-05, 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(0.0005, 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(5.6016e-06, 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(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6770e-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(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7208e-05, device='cuda:0', grad_fn=<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(5.4805e-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(3.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-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(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(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(5.3737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4427e-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(9.0837e-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(2.5767e-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(7.9794e-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(0.0001, 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(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1419e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-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(6.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9660e-05, device='cuda:0', grad_fn=<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(7.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7094e-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(0.0001, 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(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8186e-05, 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(9.5572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-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(3.3426e-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(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4557e-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(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, 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(4.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6097e-05, 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(3.8854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, 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(6.6008e-06, 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(3.3336e-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(5.0570e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7196e-05, 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(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6064e-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(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2556e-05, device='cuda:0', grad_fn=<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(9.3302e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2451e-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(7.5279e-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(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6392e-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(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0373e-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(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, 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(9.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6234e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6552e-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(7.9331e-06, 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(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3901e-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(5.4199e-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(5.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5225e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0683e-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(8.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6639e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4537e-05, device='cuda:0', grad_fn=<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(8.5387e-06, 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(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9107e-05, device='cuda:0', grad_fn=<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(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4070e-05, device='cuda:0', grad_fn=<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(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9770e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-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(1.8470e-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(9.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3294e-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(3.3033e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4565e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3779e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-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(9.7451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7898e-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(3.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-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(6.7964e-05, 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(4.7240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4695e-05, device='cuda:0', grad_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(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-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(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(2.6614e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8625e-05, device='cuda:0', grad_fn=<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(2.6433e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7077e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6224e-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(0.0003, 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(5.4805e-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(7.1687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9855e-05, 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(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-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(7.8725e-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(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2659e-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(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2038e-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(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, 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(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9937e-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(1.0719e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4228e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-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(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2019e-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(0.0002, 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(6.5845e-05, device='cuda:0', grad_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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5347e-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(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8230e-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(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, 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(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9310e-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(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(8.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7028e-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(5.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2708e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6973e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2592e-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(2.4707e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-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(2.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9873e-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(6.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(6.6822e-05, device='cuda:0', grad_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(3.6464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4958e-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(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-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(7.5444e-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(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-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(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1674e-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(6.1466e-06, 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(2.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.3287e-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(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(1.5866e-05, device='cuda:0', grad_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(7.4217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6635e-05, 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(0.0003, 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(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9746e-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(0.0001, 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(0.0005, 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(4.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0925e-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(4.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8919e-05, device='cuda:0', grad_fn=<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(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6151e-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(0.0007, 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(7.5765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, 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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, 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(0.0001, 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(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, 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(4.7909e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9604e-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(1.0749e-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(6.3581e-05, device='cuda:0', grad_fn=<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(3.1338e-05, device='cuda:0', grad_fn=<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(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, 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(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(0.0002, 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(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9631e-05, device='cuda:0', grad_fn=<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(1.0900e-05, 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(5.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2864e-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(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-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(9.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8796e-05, device='cuda:0', grad_fn=<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(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2527e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6091e-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(0.0005, 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(8.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-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(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6120e-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(4.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0127e-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
				 {'precision': {'support': 0.8773333333333333, 'agreement': 0.6125, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.45901639344262296, 'partial': 0.10204081632653061}, 'recall': {'support': 0.9267605633802817, 'agreement': 0.4666666666666667, 'direct_attack': 0.09230769230769231, 'undercutter_attack': 0.56, 'partial': 0.5555555555555556}, 'f1': {'support': 0.9013698630136987, 'agreement': 0.5297297297297299, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.5045045045045045, 'partial': 0.1724137931034483}, 'support': {'support': 355, 'agreement': 105, 'direct_attack': 65, 'undercutter_attack': 50, 'partial': 9}, 'micro_avg': {'precision': 0.714041095890411, 'recall': 0.714041095890411, 'f1': 0.7140410958904111, 'support': None}, 'macro_avg': {'precision': 0.4733360033573395, 'recall': 0.5202580955820393, 'f1': 0.4501750066417049, 'support': None}, 'weighted_avg': {'precision': 0.7194543427086935, 'recall': 0.714041095890411, 'f1': 0.7049153880828001, '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
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.09818181818181818, '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': 0.17880794701986755, 'partial': 0.0}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.09818181818181818, 'recall': 0.09818181818181818, 'f1': 0.09818181818181818, 'support': None}, 'macro_avg': {'precision': 0.019636363636363636, 'recall': 0.2, 'f1': 0.03576158940397351, 'support': None}, 'weighted_avg': {'precision': 0.009639669421487604, 'recall': 0.09818181818181818, 'f1': 0.01755568934376881, 'support': None}}
Loss: tensor(1.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5847, 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
				 {'precision': {'support': 0.8363636363636363, 'agreement': 0.6477272727272727, 'direct_attack': 0.22727272727272727, 'undercutter_attack': 0.19642857142857142, 'partial': 0.11009174311926606}, 'recall': {'support': 0.7033639143730887, 'agreement': 0.5377358490566038, 'direct_attack': 0.1282051282051282, 'undercutter_attack': 0.2037037037037037, 'partial': 0.5}, 'f1': {'support': 0.7641196013289036, 'agreement': 0.5876288659793815, 'direct_attack': 0.16393442622950816, 'undercutter_attack': 0.19999999999999998, 'partial': 0.1804511278195489}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.5727272727272728, 'recall': 0.5727272727272728, 'f1': 0.5727272727272728, 'support': None}, 'macro_avg': {'precision': 0.4035767901822947, 'recall': 0.4146017190677049, 'f1': 0.37922680427146843, 'support': None}, 'weighted_avg': {'precision': 0.6622963291920755, 'recall': 0.5727272727272728, 'f1': 0.606690980216338, 'support': None}}
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.8149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, 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(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6589, 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(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, 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(6.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3083, 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(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2811, 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
				 {'precision': {'support': 0.8896797153024911, 'agreement': 0.5214285714285715, 'direct_attack': 0.3111111111111111, 'undercutter_attack': 0.5510204081632653, 'partial': 0.2571428571428571}, 'recall': {'support': 0.764525993883792, 'agreement': 0.6886792452830188, 'direct_attack': 0.358974358974359, 'undercutter_attack': 0.5, 'partial': 0.375}, 'f1': {'support': 0.8223684210526316, 'agreement': 0.5934959349593497, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.5242718446601942, 'partial': 0.30508474576271183}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6781818181818182, 'recall': 0.6781818181818182, 'f1': 0.6781818181818182, 'support': None}, 'macro_avg': {'precision': 0.5060765326296591, 'recall': 0.537435919628234, 'f1': 0.5157108559536441, 'support': None}, 'weighted_avg': {'precision': 0.7168301080380389, 'recall': 0.6781818181818182, 'f1': 0.6917413750906494, 'support': None}}
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4363, 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(0.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2169, 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(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, 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(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, 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(0.5579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, 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(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5482, 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(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7570, 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(1.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7481, 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(0.2771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, 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(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8481, 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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1911, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, 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(0.3238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1151, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, 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(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2200, 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(0.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, 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(6.1436, 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(0.9194, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5030, 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(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.4745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.5360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, 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(2.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2948, 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(0.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, 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(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3122, 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(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5005, 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(1.6414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2556, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2049, 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(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, 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(0.4105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, 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
				 {'precision': {'support': 0.840625, 'agreement': 0.6049382716049383, 'direct_attack': 0.47058823529411764, 'undercutter_attack': 0.3655913978494624, 'partial': 0.3076923076923077}, 'recall': {'support': 0.8226299694189603, 'agreement': 0.46226415094339623, 'direct_attack': 0.20512820512820512, 'undercutter_attack': 0.6296296296296297, 'partial': 0.5}, 'f1': {'support': 0.8315301391035548, 'agreement': 0.5240641711229946, 'direct_attack': 0.28571428571428564, 'undercutter_attack': 0.46258503401360546, 'partial': 0.380952380952381}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6763636363636364, 'recall': 0.6763636363636364, 'f1': 0.6763636363636364, 'support': None}, 'macro_avg': {'precision': 0.5178870424881652, 'recall': 0.5239303910240383, 'f1': 0.49696920218136426, 'support': None}, 'weighted_avg': {'precision': 0.6990678615183279, 'recall': 0.6763636363636364, 'f1': 0.6776844795424525, 'support': None}}
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2115, 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(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, 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(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, 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(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0113, 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(0.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, 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(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7277, 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(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3064, 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(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, 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(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, 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(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0974, 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(0.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.4703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, 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(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, 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(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0172, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, 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(0.0029, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, 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(0.7495, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3447, 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(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, 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(0.0073, 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(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8201, 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(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, 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(2.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, 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(0.0014, 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(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5855, 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(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, 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(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6786, 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(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, 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(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, 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(0.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1413, 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(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, 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(0.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, 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(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, 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(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, 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(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, 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(0.1923, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, 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(0.0506, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0269, 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(0.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0159, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6010, 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(0.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, 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(0.0010, 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(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, 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(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, 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(0.0047, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0431, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, 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(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, 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(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, 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(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, 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(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6547, 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(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5122, 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(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, 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(1.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3837, 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(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, 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(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, 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(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, 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(0.3433, 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(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, 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(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1417, 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(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, 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(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, 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(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(0.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2487, 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(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, 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(0.5270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, 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(0.0068, 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(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, 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(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, 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(0.6775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, 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(0.0846, 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(0.5579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, 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(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4342, 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(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, 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(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7795, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0008, 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(0.4376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5922, 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(0.5987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5654, 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(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, 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(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2800, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, 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
				 {'precision': {'support': 0.8517441860465116, 'agreement': 0.7066666666666667, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.3888888888888889, 'partial': 0.23076923076923078}, 'recall': {'support': 0.8960244648318043, 'agreement': 0.5, 'direct_attack': 0.5897435897435898, 'undercutter_attack': 0.25925925925925924, 'partial': 0.25}, 'f1': {'support': 0.8733233979135618, 'agreement': 0.5856353591160222, 'direct_attack': 0.4259259259259259, 'undercutter_attack': 0.3111111111111111, 'partial': 0.24000000000000002}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.7072727272727273, 'recall': 0.7072727272727273, 'f1': 0.7072727272727273, 'support': None}, 'macro_avg': {'precision': 0.5022804611409263, 'recall': 0.49900546276693064, 'f1': 0.4871991588133242, 'support': None}, 'weighted_avg': {'precision': 0.7144826855315227, 'recall': 0.7072727272727273, 'f1': 0.7033185641729893, 'support': None}}
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8267, 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(0.6168, 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(0.0009, 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(0.0078, 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(0.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, 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(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, 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(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1981, 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(0.0181, 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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, 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(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7238, 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(0.0176, 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(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2938, 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(0.0097, 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(0.0733, 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(0.0247, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0350, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, 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(0.1957, 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(0.0101, 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(0.0032, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, 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(0.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.0738, 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(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, 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(0.0285, 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(0.0008, 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(0.0041, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0055, 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(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, 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(0.0138, 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(0.0038, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, 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(0.0146, 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(0.0103, 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(0.0035, 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(0.0008, 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(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, 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(0.0762, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, 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(0.0030, 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(0.0432, 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(0.0029, 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(0.0169, 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(0.0210, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, 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(0.0012, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0466, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, 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(0.0162, 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(3.5218, 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(0.0016, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0035, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2977, 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(0.0008, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, 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(0.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0028, 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(1.0510, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, 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(0.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, 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(0.0125, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(0.0180, 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(0.0043, 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(0.0220, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0287, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, 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(0.0027, 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(0.0177, 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(0.0017, 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(0.0310, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, 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(0.0029, 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(0.0018, 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(0.0399, 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(0.0008, 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(0.0027, 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(0.0053, 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(0.0013, 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(0.0013, 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(0.0697, 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(0.0091, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, 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(0.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, 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(0.0013, 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(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, 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(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0041, 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(0.0120, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, 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(0.0621, 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(0.4596, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, 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(0.0004, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, 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(0.0018, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, 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(0.0011, 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(0.0020, 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(0.0002, 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(0.0123, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(0.0010, 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(0.1187, 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(0.0009, 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(0.0004, 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(0.0217, 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(0.0030, 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(0.4202, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9309, 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(0.3689, 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(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.0022, 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(0.0670, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, 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(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, 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(0.0006, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, 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(0.0029, device='cuda:0', grad_fn=<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(0.0018, 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(0.0186, 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(0.3351, 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(0.0064, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, 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(0.0003, 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(0.0068, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, 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(0.0349, 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(0.0363, 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(0.0002, 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(0.0012, 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(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3183, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, 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(0.0006, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, 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(0.0020, 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(0.0173, 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(0.0012, 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(0.0079, 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(0.0030, 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(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6032, 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(0.0770, 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(0.0082, 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(0.0940, 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(0.0011, 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(0.0005, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, 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(0.0044, 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(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6774, 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(0.0427, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0004, 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(0.0009, 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(0.0066, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, 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(0.0101, 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(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0432, 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(0.0023, 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(0.0086, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, 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(0.0855, 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(0.0029, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2552, 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(0.3337, 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(0.4630, 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(0.0083, 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(0.0024, 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(0.0486, 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(0.0002, 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(0.0004, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, 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(0.0730, 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(0.0009, 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(0.0002, 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(0.0124, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, 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(0.0099, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, 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(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1509, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0031, 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(0.0379, 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(0.0064, 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(0.0106, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0007, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, 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(0.0020, 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(0.0030, 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(0.0007, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(0.0064, 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(0.9967, 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(0.0196, 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(0.0742, 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(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, 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(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, 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(0.0029, 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(0.0036, 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(0.0004, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(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(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(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(0.0115, 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(0.0026, 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(0.0012, 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(0.0008, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0055, 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(0.0014, 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(0.0087, device='cuda:0', grad_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(0.0577, 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(0.0024, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5898, 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(0.0016, 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(0.0003, 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(0.0198, 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(0.0007, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0038, 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(0.0024, 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(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2345, 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(0.0034, 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(0.0003, 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(0.0030, 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(0.0006, 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(0.0003, 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(1.9716, 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(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, 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(0.0035, 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(0.0026, 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(0.0015, 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(0.0038, 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(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0017, 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(0.0035, 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(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2321, 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(0.0034, 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(0.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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
				 {'precision': {'support': 0.849624060150376, 'agreement': 0.5230769230769231, 'direct_attack': 0.3, 'undercutter_attack': 0.2702702702702703, 'partial': 0.3}, 'recall': {'support': 0.691131498470948, 'agreement': 0.6415094339622641, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.37037037037037035, 'partial': 0.25}, 'f1': {'support': 0.7622259696458685, 'agreement': 0.5762711864406779, 'direct_attack': 0.3636363636363637, 'undercutter_attack': 0.3125, 'partial': 0.2727272727272727}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6145454545454545, 'recall': 0.6145454545454545, 'f1': 0.6145454545454545, 'support': None}, 'macro_avg': {'precision': 0.44859425069951386, 'recall': 0.48290995286840876, 'f1': 0.4574721584900365, 'support': None}, 'weighted_avg': {'precision': 0.6668505747453115, 'recall': 0.6145454545454545, 'f1': 0.6326089282985157, 'support': None}}
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, 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(0.0012, 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(0.0062, 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(0.0038, 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(0.0053, 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(0.0367, 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(0.0030, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, 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(0.0014, 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(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, 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(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, 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(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, 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(0.0225, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0115, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, 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(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, 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(0.0241, 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(0.0031, 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(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0052, 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(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0059, 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(0.0139, 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(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0008, 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(0.0019, 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(0.0007, 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(0.0002, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, 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(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, 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(0.0013, 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(0.0020, 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(0.0028, 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(0.0565, 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(0.0015, 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(0.0160, 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(0.0264, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, 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(0.0019, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3820, 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(0.0027, 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(0.0010, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0272, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0014, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, 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(0.0014, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, 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(0.0148, 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(0.0070, 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(0.1484, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0116, 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(0.0117, 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(0.0096, 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(0.0023, 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(0.0049, 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(0.0011, 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(0.0074, 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(0.0234, 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(0.0025, 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(0.0001, 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(0.0017, 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(0.0015, 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(0.0006, 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(0.0407, 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(0.0024, 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(0.0221, 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(0.0007, 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(0.0092, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, 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(0.0047, 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(0.0111, 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(0.0069, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3309, 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(0.0014, 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(0.0020, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, 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(0.0002, 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(0.0048, device='cuda:0', grad_fn=<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(0.0201, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(0.0041, 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(0.0013, 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(0.0042, 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(0.0005, 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(0.0014, 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(0.0772, 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(0.0005, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, 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(0.0030, 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(0.0004, 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(0.0049, 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(0.0130, 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(0.0055, 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(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, 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(0.0010, 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(0.0169, 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(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, 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(0.0048, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0003, 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(0.0025, 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(0.0145, 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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, 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(0.0125, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, 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(0.0070, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<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(0.0331, 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(0.0018, 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(0.0013, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5120, 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(0.0033, 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(0.0045, 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(0.0003, 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(0.0013, 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(0.0003, 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(0.0008, 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(0.0151, 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(0.0531, device='cuda:0', grad_fn=<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(1.1394, 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(0.0317, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0562, 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(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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0036, 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(0.0003, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9692, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, 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(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0015, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4699, 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(0.0013, 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(0.0004, 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(0.0072, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0007, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8434, 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(0.0007, 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(0.0024, 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(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(0.0040, 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(0.0019, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4811, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2024, 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(0.0015, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, 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(2.1084, 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(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7053, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831, 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(0.0487, 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(0.0039, 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(0.0051, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.1735, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, 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(0.0160, 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(0.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2300, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, 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(0.2095, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0002, 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(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<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(0.0070, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1041, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(0.0025, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, 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(0.0224, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7207, device='cuda:0', grad_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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, 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(0.0033, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, 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(0.0130, 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(0.0032, 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(0.0069, 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(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, 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(8.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, 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(0.1184, device='cuda:0', grad_fn=<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(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0485, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8296e-05, 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(0.0229, 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(0.0269, 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(0.0034, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.5176, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, 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(0.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<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(4.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9669e-05, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, 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(0.0055, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1907, 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(0.2106, 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(0.0088, 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(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1524, 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(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0058, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9424, 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(0.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, 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(0.0839, 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(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5972, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, 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(0.0067, 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(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0302, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, 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(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, 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(0.0069, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, 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(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0062, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2449, device='cuda:0', grad_fn=<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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<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(0.0197, 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(0.0048, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(8.7984e-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(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<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(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, 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(0.0446, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, 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(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2275, 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(6.3643e-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(0.0013, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, 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(0.0036, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, 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(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8231, 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(0.1772, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, 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(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, 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(0.0198, 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(0.0008, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0923, 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(0.0035, 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(0.0306, 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(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, 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(0.0024, 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(0.0066, 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(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4309e-05, 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(0.0036, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(8.0718e-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(0.0070, 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(0.0081, 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(0.0167, 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(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, 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
				 {'precision': {'support': 0.8327868852459016, 'agreement': 0.5544554455445545, 'direct_attack': 0.27522935779816515, 'undercutter_attack': 0.5217391304347826, 'partial': 0.5}, 'recall': {'support': 0.7767584097859327, 'agreement': 0.5283018867924528, 'direct_attack': 0.7692307692307693, 'undercutter_attack': 0.2222222222222222, 'partial': 0.25}, 'f1': {'support': 0.8037974683544303, 'agreement': 0.5410628019323672, 'direct_attack': 0.40540540540540543, 'undercutter_attack': 0.3116883116883117, 'partial': 0.3333333333333333}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6509090909090909, 'recall': 0.6509090909090909, 'f1': 0.6509090909090909, 'support': None}, 'macro_avg': {'precision': 0.5368421638046807, 'recall': 0.5093026576062754, 'f1': 0.4790574641427696, 'support': None}, 'weighted_avg': {'precision': 0.6945480849104351, 'recall': 0.6509090909090909, 'f1': 0.6560661978158352, 'support': None}}
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1003, 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(0.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6732e-05, device='cuda:0', grad_fn=<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(0.0032, 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(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(0.0266, 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(0.0328, 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(0.0002, 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(0.0025, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, 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(0.1459, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, 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(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3402e-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(0.0027, 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(0.6485, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, 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(0.0002, 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(0.0327, 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(0.0026, 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(0.0304, 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(0.0009, 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(0.0094, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, 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(0.0084, 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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2836, 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(0.0004, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, 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(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, 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(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0023, 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(0.0013, 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(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0001, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0294, 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(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8403, 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(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, 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(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, 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(0.0007, 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(0.0018, 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(0.0032, 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(0.0071, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0010, 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(0.0005, 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(0.0018, 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(0.0228, 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(0.0465, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, 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(0.0135, 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(0.0221, 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(0.0136, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7090, 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(0.0005, 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(0.0015, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, 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(0.0002, 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(1.0122, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8416, 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(0.0003, 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(0.0025, 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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9205, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, 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(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0229, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0680, 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(0.0010, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5018, 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(0.0013, 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(0.0006, 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(0.0002, 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(9.4038e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7257e-05, 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(0.0035, 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(0.1135, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6489, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(0.0004, 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(0.0019, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0120, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(0.0100, 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(0.0006, 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(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9891e-05, 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(0.0089, 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(0.0001, 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(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, 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(0.0003, 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(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, 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(0.0020, 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(0.1224, device='cuda:0', grad_fn=<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(0.0022, 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(0.0027, 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(0.0113, 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(0.0062, 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(8.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1071e-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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-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(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, 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(9.2101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0037, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, 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(0.0002, 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(0.0004, 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(0.6780, 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(0.0055, 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(0.0498, device='cuda:0', grad_fn=<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(0.0036, 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(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0011, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0300, 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(0.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8700e-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(0.0009, 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(0.0140, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0026, 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(0.0047, 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(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, 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(0.0082, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1967, 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(0.0002, 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(0.0002, 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(0.0029, 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(0.0002, 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(0.0014, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0023, 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(0.0043, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, 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(0.0042, 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(1.8844, 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(2.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, 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(0.0034, 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(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(0.0033, 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(0.0027, 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(0.0394, 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(0.0199, 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(9.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, 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(0.1766, 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(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, 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(0.0196, 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(0.0003, 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(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, 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(0.0035, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(7.0182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3433e-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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.1242, device='cuda:0', grad_fn=<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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, 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(0.0089, 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(0.2425, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, 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(1.2200, 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(0.0016, 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(3.9322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6016, device='cuda:0', grad_fn=<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(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, 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(0.1142, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2530, 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(0.0071, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4229, device='cuda:0', grad_fn=<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(0.0190, 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(0.0013, 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(0.0705, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8316, 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(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4760, 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(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0475e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1899, 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(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.0076, 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(0.1079, 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(0.0762, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, 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(0.0180, 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(0.0388, 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(8.9497e-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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, 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(0.0051, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0971, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0079, 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(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4451, 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(0.0048, 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(0.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4160, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2599, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, 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(0.0514, 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(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.0016, 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(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, 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(0.0075, 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(0.0193, 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(0.0043, 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(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, 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(0.1478, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6232, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, 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(1.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, 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(0.0234, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, 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(0.0031, 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(0.0006, 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(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, 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(1.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, 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(0.0097, 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(0.0028, 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(0.0008, 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(0.0008, 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(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, 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(0.0018, 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(0.0005, 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(0.0224, 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(0.0008, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<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(0.1920, 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(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, 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(0.0014, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2171, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3127e-05, 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(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-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(0.1061, device='cuda:0', grad_fn=<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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(6.0191e-05, 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(0.0133, 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(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4451e-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(0.0003, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3713, 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(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, 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(0.0365, 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(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1081e-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(1.0241, 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(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4230e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3965, 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(0.0008, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.1372, 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>)
			------------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
				 {'precision': {'support': 0.8511326860841424, 'agreement': 0.5197368421052632, 'direct_attack': 0.41379310344827586, 'undercutter_attack': 0.42, 'partial': 0.4}, 'recall': {'support': 0.8042813455657493, 'agreement': 0.7452830188679245, 'direct_attack': 0.3076923076923077, 'undercutter_attack': 0.3888888888888889, 'partial': 0.16666666666666666}, 'f1': {'support': 0.8270440251572327, 'agreement': 0.6124031007751939, 'direct_attack': 0.35294117647058826, 'undercutter_attack': 0.40384615384615385, 'partial': 0.23529411764705882}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6890909090909091, 'recall': 0.6890909090909091, 'f1': 0.6890909090909091, 'support': None}, 'macro_avg': {'precision': 0.5209325263275362, 'recall': 0.4825624455363074, 'f1': 0.4863057147792455, 'support': None}, 'weighted_avg': {'precision': 0.6942371357221004, 'recall': 0.6890909090909091, 'f1': 0.6846865125857461, 'support': None}}
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(5.7285e-05, device='cuda:0', grad_fn=<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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3880, 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(0.0005, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7488, 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(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, 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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, 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(0.0025, 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(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, 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(0.0002, 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(0.0015, 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(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, 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(1.1613, 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(0.0008, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7793, device='cuda:0', grad_fn=<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(0.0252, 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(0.0020, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2773, 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(0.0004, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0285, 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(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0023, 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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, 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(0.0076, 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(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, 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(1.1967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0324, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6773e-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(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, 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(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0132, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0001, 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(9.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, 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(0.1839, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0009, 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(0.0008, 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(1.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<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(5.6831e-05, 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(0.0033, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, 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(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, 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(0.0044, 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(0.3170, 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(4.6182, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(0.0157, 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(1.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, 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(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2010e-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(9.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, 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(0.0011, 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(0.0162, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.4165, 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(0.0147, 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(0.0345, 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(0.0014, 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(2.3055, 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(8.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-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(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0081, 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(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(1.0578, 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(0.0837, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, 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(0.0621, 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(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6453, 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(0.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, 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(0.0004, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<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(0.9576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0022, 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(0.0010, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, 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(0.0166, 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(0.0211, 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(0.0031, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0024, 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(0.0238, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, 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(9.6914e-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(0.0145, device='cuda:0', grad_fn=<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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, 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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(8.1958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7055, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, 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(0.0231, 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(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6288, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, 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(9.1677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.2991, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, 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(0.0010, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, 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(0.0003, 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(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.0026, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, 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(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, 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(1.4433, 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(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3779, 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(0.0002, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, 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(0.0024, 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(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, 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(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, 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(0.0005, 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(0.0015, 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(0.0005, 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(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1024, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, 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(0.0067, 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(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, 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(0.0001, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0126, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, 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(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, 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(0.0033, device='cuda:0', grad_fn=<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(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, 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(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, 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(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, 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(0.0029, 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(0.0045, 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(0.2540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1569, 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(0.0035, 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(0.0048, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, 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(0.0011, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, 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(0.5405, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, 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(0.2063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0219, 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(0.0608, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, 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(0.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5036e-05, device='cuda:0', grad_fn=<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(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, 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(0.0078, 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(0.0023, 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(8.1564e-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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2852, device='cuda:0', grad_fn=<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(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, 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(0.0004, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, 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(0.0002, 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(0.0107, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, 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(0.0068, 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(0.0015, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, 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(0.0005, 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(0.0044, 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(8.2504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, 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(1.6361, device='cuda:0', grad_fn=<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(0.0025, 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(0.0819, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, 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(0.0041, 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(0.0042, 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(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6627, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4582e-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(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, 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(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0016, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4143, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-05, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.0002, 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(0.0045, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6147, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5824e-05, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6732e-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(0.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1000e-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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0496e-05, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, 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(0.0262, 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(0.0003, 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(0.0003, 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(0.0005, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, 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(0.0068, 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(0.0003, 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(0.0061, 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(0.0045, 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(0.0014, 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(0.0040, 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(0.0006, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, 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(0.0267, 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(0.0187, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, 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(0.2507, 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(0.0091, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, 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(0.0013, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-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(0.0017, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<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(0.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, 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(0.0019, device='cuda:0', grad_fn=<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(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6686, 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(4.8444e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, 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(0.0206, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(7.8871e-05, 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(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, 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(8.0686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1995, 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(0.6253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4795e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0003, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, 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(0.0109, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0945, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8477e-05, 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(0.0016, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, 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(0.0001, 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(0.0014, 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(8.1747e-05, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(7.8962e-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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4053, 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(0.0006, 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(8.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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
				 {'precision': {'support': 0.7927170868347339, 'agreement': 0.5233644859813084, 'direct_attack': 0.3448275862068966, 'undercutter_attack': 0.4375, 'partial': 0.6666666666666666}, 'recall': {'support': 0.8654434250764526, 'agreement': 0.5283018867924528, 'direct_attack': 0.2564102564102564, 'undercutter_attack': 0.3888888888888889, 'partial': 0.25}, 'f1': {'support': 0.827485380116959, 'agreement': 0.5258215962441314, 'direct_attack': 0.29411764705882354, 'undercutter_attack': 0.411764705882353, 'partial': 0.36363636363636365}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6836363636363636, 'recall': 0.6836363636363636, 'f1': 0.6836363636363636, 'support': None}, 'macro_avg': {'precision': 0.5530151651379211, 'recall': 0.45780889143361014, 'f1': 0.4845651385877262, 'support': None}, 'weighted_avg': {'precision': 0.6686698159473556, 'recall': 0.6836363636363636, 'f1': 0.6704690246915227, 'support': None}}
Loss: tensor(0.0041, 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(0.1009, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, 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(0.0229, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4491e-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(0.0191, device='cuda:0', grad_fn=<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(7.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, 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(0.0012, 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(0.2428, device='cuda:0', grad_fn=<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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(0.0031, 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(0.0001, 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(0.0139, 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(0.0001, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1689, 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(8.9525e-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(0.0807, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9375e-05, 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(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(0.0107, 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(0.0017, 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(0.0004, 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(0.0011, 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(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6480, 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(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1737e-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(0.0023, 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(0.0012, 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(0.0069, 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(0.0014, 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(0.0014, 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(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, 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(6.4854e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, 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(0.0039, 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(9.9336e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0033, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3312e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0354e-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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, 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(0.0002, 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(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, 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(0.2493, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, 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(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-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(0.0004, 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(0.0789, 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(7.3482e-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(0.0039, 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(0.0015, 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(0.0031, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5609e-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(0.0502, 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(8.7802e-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(0.0011, 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(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-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(9.3008e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7396e-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(0.0401, 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(0.0017, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0015, 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(0.0013, 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(0.0001, 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(0.0098, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1140e-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(0.0004, 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(9.6365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, 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(0.0155, 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(0.0002, 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(0.0008, 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(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2079e-05, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-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(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9911e-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(0.0007, 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(0.0009, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4204, 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(0.0021, 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(0.0059, 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(5.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-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(9.0405e-05, 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(0.0025, 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(0.0020, 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(0.0018, 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(0.0044, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, 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(0.1317, 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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, 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(0.0105, 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(0.0004, 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(0.0125, device='cuda:0', grad_fn=<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(1.4641, device='cuda:0', grad_fn=<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(0.0945, 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(0.0023, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8643, device='cuda:0', grad_fn=<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(0.0209, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0869e-05, 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(0.0008, 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(0.0021, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, 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(0.0070, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, 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(0.0004, 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(0.0444, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0001, 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(0.0021, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, 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(0.0054, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9830e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3994, device='cuda:0', grad_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(0.0097, device='cuda:0', grad_fn=<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(0.3192, 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(6.5913e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(0.0002, 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(9.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-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(0.0007, 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(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-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(0.0001, 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(0.0002, 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(0.0285, device='cuda:0', grad_fn=<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(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, 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(0.0020, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, 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(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3997e-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(8.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, 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(0.0019, 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(0.0001, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0061e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9043e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9858e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3554, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-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(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<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(4.7657e-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(0.0315, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8333, 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(0.0004, 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(0.0003, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-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(0.0009, 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(0.0032, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5419e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6368e-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(0.0123, 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(0.0725, device='cuda:0', grad_fn=<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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-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(0.0078, 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(0.0046, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-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(0.1203, 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(4.8050e-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(0.0001, 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(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(8.7105e-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(0.0027, 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(0.0065, device='cuda:0', grad_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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3259e-05, 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(0.0025, 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(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1010e-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(0.0008, 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(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9376e-05, 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(0.0041, 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(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(0.0019, 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(0.0069, 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(8.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(9.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9699e-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(0.0011, 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(5.6649e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.4954e-05, device='cuda:0', grad_fn=<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(0.0021, 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(6.4944e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8851e-05, 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(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-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(7.8083e-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(0.0002, 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(0.0124, 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(0.0016, 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(9.6428e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, 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(0.0058, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, 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(7.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, 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(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-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(0.0004, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9215e-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(9.8579e-05, device='cuda:0', grad_fn=<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(0.0758, 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(7.2179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-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(0.0011, 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(0.0002, 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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, 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(7.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3081, device='cuda:0', grad_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(0.0027, 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(0.0015, 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(0.0024, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4714e-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(0.0002, 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(0.0006, 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(0.0068, 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(0.0004, 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(0.0002, 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(0.0051, 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(0.0014, 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(0.0008, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-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(0.0044, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, 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(0.0004, 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(0.0002, 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(0.0007, 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(0.0031, 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(0.0038, 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(0.0007, 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(0.0097, 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(0.0210, device='cuda:0', grad_fn=<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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0021, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-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(0.0009, 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(0.0209, 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(0.0024, 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(0.0008, 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(0.0008, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3070e-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(0.0679, 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(0.0058, 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(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8971e-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(0.0002, 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(0.0029, 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(0.0016, 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(8.9164e-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(0.0024, 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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0003, 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(0.0005, 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(0.0010, 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(0.1616, 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(0.0005, 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(0.0079, 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(0.0001, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3118e-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(0.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3675e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6618e-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(0.0005, 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(0.0025, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-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(0.0029, 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(4.2328e-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(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2353e-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(0.0002, 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(5.8768e-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(0.0003, 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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(5.4560e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, 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(0.0003, 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(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(0.0009, 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(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-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(0.0015, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6146e-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(0.0011, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2987e-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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-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(0.0102, 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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0010e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0007, 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(0.0004, 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(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-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(6.4309e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4118e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4968, 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(0.0059, 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(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7760e-05, 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(9.6520e-05, 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(0.0081, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4886e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4364, device='cuda:0', grad_fn=<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(0.0014, 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(0.0004, 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(0.0019, 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(0.0154, 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(8.2110e-05, 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
				 {'precision': {'support': 0.8287461773700305, 'agreement': 0.52, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.42105263157894735, 'partial': 0.47368421052631576}, 'recall': {'support': 0.8287461773700305, 'agreement': 0.6132075471698113, 'direct_attack': 0.5128205128205128, 'undercutter_attack': 0.14814814814814814, 'partial': 0.375}, 'f1': {'support': 0.8287461773700304, 'agreement': 0.5627705627705627, 'direct_attack': 0.40404040404040403, 'undercutter_attack': 0.2191780821917808, 'partial': 0.4186046511627907}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6781818181818182, 'recall': 0.6781818181818182, 'f1': 0.6781818181818182, 'support': None}, 'macro_avg': {'precision': 0.5153632705617254, 'recall': 0.49558447710170056, 'f1': 0.48666797550711366, 'support': None}, 'weighted_avg': {'precision': 0.6785913875598086, 'recall': 0.6781818181818182, 'f1': 0.6696243335954881, 'support': None}}
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-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(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(0.0015, 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(0.0008, 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(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(0.0029, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<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(0.0004, 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(7.7993e-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(0.0003, 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(0.0009, 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(0.0021, 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(0.0007, 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(0.0004, 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(0.2280, device='cuda:0', grad_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(0.0005, 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(0.0003, 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(0.0002, 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(0.0017, 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(0.0023, 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(0.0006, 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(0.0009, 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(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(0.0012, 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(0.0003, 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(0.0064, 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(0.0048, 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(0.0045, 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(0.0006, 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(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3009e-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(0.0015, 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(0.0043, 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(0.0002, 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(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(0.0366, 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(5.9858e-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(0.0019, 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(0.0126, 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(0.0002, 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(0.0002, 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(0.0006, 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(0.0002, 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(0.0027, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5823e-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(0.0004, 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(0.0002, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0002, 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(0.0020, 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(0.0009, 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(0.0009, 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(0.0002, 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(0.0007, 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(0.0003, 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(0.0013, 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(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(0.0024, 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(0.0031, 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(0.0001, 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(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(0.0002, 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(0.0019, 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(0.7776, device='cuda:0', grad_fn=<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(0.0027, 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(0.0009, 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(0.0007, 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(0.0009, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0002, 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(0.0009, 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(4.9565e-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(7.3754e-05, 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(0.0041, 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(0.0010, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-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(0.0001, 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(0.0032, 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(0.0132, 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(0.0023, 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(0.0007, 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(8.3653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4278e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2615e-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(0.0007, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0010, 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(8.7257e-05, 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(7.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, 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(0.0005, 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(0.0010, 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(7.6479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6116e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4187e-05, 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(0.0009, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9637e-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(0.0036, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6218e-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(0.0034, 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(5.5923e-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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-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(0.0028, 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(0.0027, 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(0.0003, 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(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6426e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-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(0.0013, 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(0.0003, 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(6.4854e-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(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, 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(6.7912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(3.1974e-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(0.0032, 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(5.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(7.6146e-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(0.0030, 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(0.0065, 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(0.0010, 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(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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1896e-05, 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(5.4136e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-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(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(8.2595e-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(0.0013, 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(0.0071, device='cuda:0', grad_fn=<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(0.0013, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(6.7851e-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(8.2232e-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(0.0005, 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(8.8496e-05, device='cuda:0', grad_fn=<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(7.5388e-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(9.7275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(0.0012, 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(0.0002, 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(0.0010, 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(0.0008, 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(7.5904e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<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(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(0.0007, 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(0.0023, 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(0.0037, 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(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-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(7.0242e-05, 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(0.0008, 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(0.0010, 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(0.0058, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, 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(8.4074e-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(0.0017, 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(4.7627e-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(0.0005, 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(0.0004, 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(6.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8092e-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(4.8686e-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(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-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(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1040e-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(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-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(0.0025, 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(0.0002, 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(0.0013, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-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(0.0010, 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(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4963e-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(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1947e-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(0.0008, 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(0.0011, 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(0.0017, 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(5.1653e-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(0.0004, 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(0.0019, 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(0.0018, 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(0.0003, 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(0.0445, 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(4.3388e-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(0.0112, 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(7.3936e-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(0.0021, 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(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-05, 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(0.0004, 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(0.0050, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-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(0.0003, 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(2.9067e-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(0.0006, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, 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(0.0005, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4138e-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(0.0009, 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(4.6779e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4642e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-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(0.0007, 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(0.0021, 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(0.0006, 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(0.0007, 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(0.0010, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-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(0.0038, 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(0.0005, 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(0.0009, 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(0.0004, 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(7.1756e-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(0.0001, 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(4.8717e-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(0.0013, 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(0.0015, device='cuda:0', grad_fn=<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(0.0004, 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(0.0003, 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(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0037, device='cuda:0', grad_fn=<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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0037, 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(0.0014, 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(0.0002, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8959e-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(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1493e-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(7.0726e-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(9.0161e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-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(0.0011, 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(9.3248e-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(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6165e-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(0.0014, 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(0.0002, 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(7.3632e-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(8.3653e-05, 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(0.0042, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6317e-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(5.4802e-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(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4874e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4490e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0006, 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(0.0005, 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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1482e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-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(0.0032, 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(0.0015, 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(8.1656e-05, 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(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9253e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-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(6.9486e-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(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9877e-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(2.6130e-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(4.8929e-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(0.0007, 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(0.0008, 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(9.4008e-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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1442e-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(9.9546e-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(0.0015, 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(0.0008, 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(0.0109, 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(8.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-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(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(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, 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(0.0012, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-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(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-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(0.0027, 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(3.9513e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3289e-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(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(0.0002, 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(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1878e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-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(0.0017, 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(0.0216, 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(3.7151e-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(0.0005, 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(4.5719e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(4.6143e-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(0.0001, 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(0.0005, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-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(6.1796e-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(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3380e-05, 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(0.0033, 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(7.1573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3702e-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(0.0008, 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(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-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(8.2050e-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(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0011, 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(3.4578e-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(9.5340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3611e-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(0.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9726e-05, 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(0.0012, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5954e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-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(5.3409e-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(0.0029, 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(9.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<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(0.0004, 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(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(6.2008e-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(0.0015, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.782258064516129, 'agreement': 0.5875, 'direct_attack': 0.37209302325581395, 'undercutter_attack': 0.39473684210526316, 'partial': 0.5294117647058824}, 'recall': {'support': 0.8899082568807339, 'agreement': 0.44339622641509435, 'direct_attack': 0.41025641025641024, 'undercutter_attack': 0.2777777777777778, 'partial': 0.375}, 'f1': {'support': 0.8326180257510729, 'agreement': 0.5053763440860214, 'direct_attack': 0.3902439024390244, 'undercutter_attack': 0.32608695652173914, 'partial': 0.4390243902439025}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6872727272727273, 'recall': 0.6872727272727273, 'f1': 0.6872727272727273, 'support': None}, 'macro_avg': {'precision': 0.5331999389166177, 'recall': 0.4792677342660032, 'f1': 0.49866992380835207, 'support': None}, 'weighted_avg': {'precision': 0.666557612418866, 'recall': 0.6872727272727273, 'f1': 0.6712741456488522, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(5.0715e-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(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7708e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5823e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7848e-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(0.0004, 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(5.3622e-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(0.0047, 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(0.0041, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0019, 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(8.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, 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(0.0005, 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(0.0021, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3429e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-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(0.0009, 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(6.8154e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-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(0.0022, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(6.9455e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0023, 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(0.0031, 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(0.0016, 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(0.0001, 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(0.0011, 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(0.0002, 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(0.0006, 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(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3914e-05, 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(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(7.9111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9724e-05, 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(0.0005, 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(8.8163e-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(4.5326e-05, 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(4.4084e-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(5.8163e-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(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2461e-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(0.0002, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-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(0.0051, 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(0.0003, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0523e-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(4.9897e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-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(0.0037, 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(5.0624e-05, 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(0.0005, 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(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5126e-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(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-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(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9494e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-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(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-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(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0009, 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(0.0001, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7912e-05, 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(4.3327e-05, 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(0.0031, 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(0.0011, 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(0.0009, 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(8.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(3.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8474e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-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(0.0008, 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(0.0023, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-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(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(0.0001, 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(0.0020, 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(0.0004, 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(3.1731e-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(0.0009, 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(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3342e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9939e-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(0.0042, 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(0.0011, 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(0.0004, 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(3.9876e-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(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(8.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, 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(0.0032, 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(0.0008, 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(0.0192, device='cuda:0', grad_fn=<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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3520e-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(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, 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(5.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-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(3.5637e-05, device='cuda:0', grad_fn=<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(3.4275e-05, 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(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(0.0025, 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(0.0022, 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(0.0021, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-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(9.3432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2401e-05, 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(9.3160e-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(0.0023, 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(3.2337e-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(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, 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(0.0017, 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(5.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2986e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7133e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2422e-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(4.2237e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.2774e-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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-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(0.0001, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1219e-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(7.1785e-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(0.0007, 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(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(0.0013, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3077e-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(5.1956e-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(0.0004, 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(0.0001, 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(3.3669e-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(8.0140e-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(0.0034, 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(8.3837e-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(8.0807e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-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(0.0010, 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(0.0012, 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(4.7173e-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(9.4883e-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(0.0002, 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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-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(3.1005e-05, device='cuda:0', grad_fn=<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(2.8855e-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(4.1390e-05, device='cuda:0', grad_fn=<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(4.8141e-05, device='cuda:0', grad_fn=<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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2612e-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(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(4.6295e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-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(3.3820e-05, device='cuda:0', grad_fn=<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(9.9081e-05, 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(0.0003, 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(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-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(0.0005, 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(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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-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(0.0012, 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(0.0013, 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(0.0024, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8876e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9624e-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(0.0012, 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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-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(0.0009, 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(0.0003, 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(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(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(0.0004, 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(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(7.3542e-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(0.0006, 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(5.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-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(9.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, 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(0.0014, 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(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7244e-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(0.0004, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8980e-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(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6006e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3765e-05, device='cuda:0', grad_fn=<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(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-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(0.0003, 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(2.9309e-05, 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(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(5.4378e-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(4.5568e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3239e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-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(6.5731e-05, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-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(4.6597e-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(0.0014, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-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(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-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(0.0007, 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(0.0002, 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(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(2.8401e-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(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1240e-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(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-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(0.0059, 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(0.0005, 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(7.1664e-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(0.0016, 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(0.0002, 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(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(9.8367e-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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-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(0.0004, 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(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-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(9.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, 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(0.0011, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-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(3.5577e-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(0.0028, 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(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(4.7172e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0777e-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(0.0014, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, 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(0.0003, 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(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8728e-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(2.8552e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-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(0.0006, 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(8.5208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, 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(6.8941e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0017, 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(0.0013, 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(0.0004, 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(0.0004, 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(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4481e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-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(5.8678e-05, 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(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(0.0005, 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(0.0001, 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(0.0003, 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(4.9746e-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(0.0043, 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(3.9876e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3451e-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(3.3003e-05, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1302e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-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(3.1883e-05, device='cuda:0', grad_fn=<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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2563e-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(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(5.9465e-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(3.7303e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-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(0.0001, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-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(4.1087e-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(2.4223e-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(0.0045, 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(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(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(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(0.0003, 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(0.0025, 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(0.0015, 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(0.0034, device='cuda:0', grad_fn=<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(0.0010, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6748e-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(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(0.0015, 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(7.2240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2159e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-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(2.6493e-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(0.0005, 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(6.0403e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7899159663865546, 'agreement': 0.5632183908045977, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.3488372093023256, 'partial': 0.5333333333333333}, 'recall': {'support': 0.8623853211009175, 'agreement': 0.46226415094339623, 'direct_attack': 0.4358974358974359, 'undercutter_attack': 0.2777777777777778, 'partial': 0.3333333333333333}, 'f1': {'support': 0.8245614035087718, 'agreement': 0.5077720207253887, 'direct_attack': 0.3908045977011494, 'undercutter_attack': 0.3092783505154639, 'partial': 0.4102564102564102}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6745454545454546, 'recall': 0.6745454545454546, 'f1': 0.6745454545454546, 'support': None}, 'macro_avg': {'precision': 0.5178943132986956, 'recall': 0.4743316038105722, 'f1': 0.4885345565414368, 'support': None}, 'weighted_avg': {'precision': 0.6608243267927568, 'recall': 0.6745454545454546, 'f1': 0.6640799585974424, '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(0.0007, 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(3.5758e-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(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-05, device='cuda:0', grad_fn=<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(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7627e-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(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-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(0.0004, 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(0.0031, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9654e-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(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7031e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0918e-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(8.0686e-05, device='cuda:0', grad_fn=<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(0.0029, 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(0.0008, 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(9.0253e-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(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7502e-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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0323e-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(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8306e-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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, 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(8.0385e-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(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-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(5.4583e-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(0.0013, 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(0.0005, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8949e-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(0.0007, 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(3.0914e-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(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9031e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-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(2.5737e-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(0.0020, 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(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(0.0014, 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(0.0004, 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(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2693e-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(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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6992e-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(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, 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(0.0002, 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(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7274e-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(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4861e-05, 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(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(3.3730e-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(6.3794e-05, 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(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8354e-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(0.0008, 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(6.6367e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-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(0.0022, 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(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4845e-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(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(4.1784e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-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(6.2980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0494e-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(0.0002, 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(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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-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(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-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(0.0006, 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(0.0020, 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(0.0001, 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(6.4822e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-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(0.0001, 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(0.0005, 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(3.5092e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7344e-05, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, 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(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-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(0.0009, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-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(9.9820e-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(3.1701e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-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(6.6639e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-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(9.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3257e-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(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, 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(0.0004, 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(4.8384e-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(2.8976e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8214e-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(2.9794e-05, device='cuda:0', grad_fn=<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(3.8180e-05, device='cuda:0', grad_fn=<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(0.0007, 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(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(6.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9565e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, 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(0.0002, 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(0.0009, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4067e-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(2.3980e-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(6.2976e-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(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<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(7.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<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(7.4267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5064e-05, 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(0.0024, 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(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(0.0007, 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(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-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(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5579e-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(0.0006, 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(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<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(7.7507e-05, 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(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6547e-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(4.3479e-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(0.0024, 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(3.9876e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4258e-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(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-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(2.9340e-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(3.7030e-05, 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(4.1026e-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(7.2542e-05, device='cuda:0', grad_fn=<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(8.9070e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-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(0.0008, 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(7.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8215e-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(6.6821e-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(9.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-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(0.0001, 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(0.0004, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2258e-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(0.0002, 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(0.0009, 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(8.4956e-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(0.0005, 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(0.0007, 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(7.6900e-05, device='cuda:0', grad_fn=<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(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2995e-05, device='cuda:0', grad_fn=<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(5.6891e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<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(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6710e-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(6.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-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(6.4975e-05, 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(0.0007, 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(7.0333e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-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(6.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5884e-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(1.9530e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-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(0.0003, 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(5.7799e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0019, 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(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5369e-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(2.6766e-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(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6567e-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(0.0040, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-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(6.6328e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6959e-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(6.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-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(2.3405e-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(7.5843e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7670e-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(4.4144e-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(0.0002, 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(0.0001, 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(0.0015, 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(0.0005, 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(0.0003, 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(8.6254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8309e-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(5.6013e-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(0.0016, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-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(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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2875e-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(8.5350e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0012, 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(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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-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(0.0013, 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(0.0007, 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(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, 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(0.0003, 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(0.0022, 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(0.0002, 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(0.0005, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-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(0.0003, 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(0.0015, 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(0.0004, 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(0.0009, 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(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5225e-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(0.0003, 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(5.7527e-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(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(0.0004, 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(0.0008, 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(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6153e-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(2.7523e-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(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-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(5.4529e-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(0.0009, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-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(0.0004, 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(3.0308e-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(4.2359e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, 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(5.8193e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0009, 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(0.0004, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1704e-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(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(2.7553e-05, 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(0.0011, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7489e-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(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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2503e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9674e-05, 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(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7798e-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(0.0007, 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(0.0019, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-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(5.5407e-05, device='cuda:0', grad_fn=<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(6.0676e-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(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(0.0001, 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(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(3.5577e-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(8.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, 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(0.0004, 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(7.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5751e-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(0.0003, 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(0.0013, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0733e-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(2.5616e-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(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7408e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4529e-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(3.4275e-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(2.3950e-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(0.0031, 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(0.0007, 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(0.0012, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-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(0.0007, 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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(2.7129e-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(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8836e-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(7.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7132e-05, device='cuda:0', grad_fn=<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(3.8059e-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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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
				 {'precision': {'support': 0.7977207977207977, 'agreement': 0.5666666666666667, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.4811320754716981, 'direct_attack': 0.4358974358974359, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8259587020648967, 'agreement': 0.5204081632653061, 'direct_attack': 0.3908045977011494, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3684210526315789}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6745454545454546, 'recall': 0.6745454545454546, 'f1': 0.6745454545454546, 'support': None}, 'macro_avg': {'precision': 0.5117959325938048, 'recall': 0.47225231749638885, 'f1': 0.48448483976624956, 'support': None}, 'weighted_avg': {'precision': 0.6638488113541305, 'recall': 0.6745454545454546, 'f1': 0.6662621024471557, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-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(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(6.4985e-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(2.5979e-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(7.9839e-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(0.0005, 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(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0002, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-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(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(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(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0694e-05, device='cuda:0', grad_fn=<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(0.0004, 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(9.9497e-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(7.4541e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(3.5486e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3450e-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(0.0005, 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(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(5.2076e-05, 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(0.0010, 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(0.0002, 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(8.8757e-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(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4370e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1229e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<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(6.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8365e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0005, 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(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6330e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-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(0.0007, 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(0.0019, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-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(0.0013, 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(0.0002, 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(0.0021, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3675e-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(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-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(3.0339e-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(0.0004, 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(3.6939e-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(7.8913e-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(4.1874e-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(2.7614e-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(9.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(9.9463e-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(3.6727e-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(2.1074e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(3.5183e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-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(2.5949e-05, device='cuda:0', grad_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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1260e-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(9.7288e-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(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(0.0003, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-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(8.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, 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(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(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0200e-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(3.1701e-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(8.0807e-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(0.0004, 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(5.5618e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6021e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(3.9179e-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(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-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(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2208e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6691e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-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(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, 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(9.9324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2750e-05, 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(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, 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(0.0003, 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(2.8189e-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(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0432e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0001, 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(0.0018, 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(9.4068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(6.1129e-05, device='cuda:0', grad_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(0.0002, 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(3.5970e-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(5.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(2.3587e-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(0.0003, 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(0.0008, 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(0.0011, 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(0.0001, 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(2.4586e-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(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, 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(6.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1959e-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(0.0012, 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(6.4514e-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(7.0515e-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(2.8007e-05, 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(0.0004, 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(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(0.0008, 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(3.4063e-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(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9039e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8100e-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(8.5614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-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(3.7393e-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(1.8621e-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(7.3118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4681e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, 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(5.5134e-05, device='cuda:0', grad_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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0006, 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(0.0007, 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(3.4608e-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(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-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(0.0020, 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(6.0462e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-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(4.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9346e-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(0.0002, 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(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, 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(0.0008, 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(7.6630e-05, device='cuda:0', grad_fn=<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(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-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(0.0007, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-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(0.0002, 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(2.4768e-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(6.3430e-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(0.0017, 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(0.0003, 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(8.5888e-05, device='cuda:0', grad_fn=<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(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-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(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6367e-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(3.8665e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0383e-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(2.2255e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8979e-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(4.0693e-05, 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(7.2089e-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(4.2510e-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(7.1179e-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(0.0013, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, 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(2.0075e-05, 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(0.0010, 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(8.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-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(4.5386e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-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(0.0003, 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(3.6000e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-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(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(2.8007e-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(6.0220e-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(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0008, 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(0.0007, 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(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(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2126e-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(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-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(7.6507e-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(0.0002, device='cuda:0', grad_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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(6.9690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-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(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-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(4.7142e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3129e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-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(0.0003, 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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-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(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, 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(8.6649e-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(2.6524e-05, 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(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(2.9854e-05, 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(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(2.3920e-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(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-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(3.0702e-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(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(0.0001, 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(1.9227e-05, device='cuda:0', grad_fn=<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(0.0018, 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(7.7840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-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(8.1146e-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(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-05, device='cuda:0', grad_fn=<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(7.0908e-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(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(8.1213e-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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-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(0.0010, 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(0.0009, 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(0.0002, 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(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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2545e-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(6.8027e-05, device='cuda:0', grad_fn=<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(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(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(0.0013, 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(0.0002, 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(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(3.5849e-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(5.1865e-05, device='cuda:0', grad_fn=<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(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-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(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-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(0.0022, 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(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-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(4.5386e-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(0.0003, 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(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7560e-05, device='cuda:0', grad_fn=<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(5.0261e-05, device='cuda:0', grad_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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<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(9.4005e-05, device='cuda:0', grad_fn=<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(5.8053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-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(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8094e-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(6.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, 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(5.3924e-05, device='cuda:0', grad_fn=<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(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0004, 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(0.0010, 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(0.0002, 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(7.6206e-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(7.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-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(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3834e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(2.1104e-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(7.5326e-05, 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(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(2.3769e-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(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3783e-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(6.7833e-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(0.0002, 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(0.0012, 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(0.0007, 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(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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, 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(0.0008, 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(2.7008e-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(6.4126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6749e-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(6.9200e-05, device='cuda:0', grad_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(0.0003, 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(8.5228e-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(5.7527e-05, device='cuda:0', grad_fn=<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(5.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8192e-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(7.1724e-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(0.0002, 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(2.1982e-05, device='cuda:0', grad_fn=<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(8.0958e-05, device='cuda:0', grad_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(0.0002, 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(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6577e-05, device='cuda:0', grad_fn=<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(2.4949e-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(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1915e-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(8.1224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-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(0.0001, 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(9.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1634e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-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(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-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(0.0019, 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(5.1552e-05, device='cuda:0', grad_fn=<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(2.3496e-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>)
			------------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
				 {'precision': {'support': 0.7971428571428572, 'agreement': 0.5666666666666667, 'direct_attack': 0.3469387755102041, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5}, 'recall': {'support': 0.8532110091743119, 'agreement': 0.4811320754716981, 'direct_attack': 0.4358974358974359, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8242245199409158, 'agreement': 0.5204081632653061, 'direct_attack': 0.3863636363636364, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3684210526315789}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6727272727272727, 'recall': 0.6727272727272727, 'f1': 0.6727272727272727, 'support': None}, 'macro_avg': {'precision': 0.5102347662469243, 'recall': 0.4716406967012817, 'f1': 0.4832498110739508, 'support': None}, 'weighted_avg': {'precision': 0.6629926762194239, 'recall': 0.6727272727272727, 'f1': 0.6649161478167833, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8123e-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(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7967e-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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-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(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5578e-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(0.0033, 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(0.0001, 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(0.0004, 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(0.0004, 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(0.0003, 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(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0306e-05, 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(0.0001, 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(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(4.8867e-05, device='cuda:0', grad_fn=<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(2.3526e-05, device='cuda:0', grad_fn=<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(2.8583e-05, device='cuda:0', grad_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(0.0004, 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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0049e-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(4.6627e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-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(0.0011, 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(2.2557e-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(5.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2715e-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(7.5227e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9732e-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(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1079e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-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(0.0007, 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(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-05, device='cuda:0', grad_fn=<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(7.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5600e-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(0.0005, 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(0.0043, 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(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5184e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, 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(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0835e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6943e-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(0.0011, 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(6.2458e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9206e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6548e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.8462e-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(0.0005, 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(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6829e-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(7.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8221e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, 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(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0002, 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(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(0.0004, 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(5.6470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9513e-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(0.0003, 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(2.3647e-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(8.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-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(0.0004, 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(0.0002, 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(4.0845e-05, 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(0.0008, 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(0.0005, 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(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(4.8565e-05, device='cuda:0', grad_fn=<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(0.0002, 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(2.1861e-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(4.1662e-05, 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(6.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<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(8.9224e-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(5.2047e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8603e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(1.8864e-05, device='cuda:0', grad_fn=<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(0.0005, 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(3.1217e-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(0.0001, 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(0.0007, 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(4.2964e-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(6.1533e-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(6.3976e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6122e-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(0.0004, 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(8.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-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(6.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-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(6.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-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(4.3944e-05, 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(9.8988e-05, device='cuda:0', grad_fn=<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(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-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(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-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(0.0010, 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(0.0007, 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(7.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-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(4.6019e-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(3.5516e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1863e-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(0.0003, 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(0.0002, 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(4.0451e-05, device='cuda:0', grad_fn=<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(4.9261e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-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(0.0004, 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(7.5539e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2563e-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(3.8725e-05, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, 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(4.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0022, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-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(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-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(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3421e-05, device='cuda:0', grad_fn=<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(1.9287e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-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(2.3617e-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(0.0002, 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(4.8807e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-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(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-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(2.0983e-05, 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(0.0007, 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(7.9116e-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(0.0002, 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(0.0003, 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(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6868e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8758e-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(0.0010, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4109e-05, device='cuda:0', grad_fn=<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(0.0002, 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(5.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5203e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-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(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(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-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(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-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(0.0016, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9753e-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(2.3345e-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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9489e-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(0.0004, 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(5.0503e-05, device='cuda:0', grad_fn=<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(8.2205e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-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(1.8894e-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(8.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-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(2.9430e-05, device='cuda:0', grad_fn=<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(2.3829e-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(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4023e-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(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9859e-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(2.2648e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0804e-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(8.4104e-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(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1860e-05, device='cuda:0', grad_fn=<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(5.0503e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4181e-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(2.7008e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3015e-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(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-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(2.1982e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4934e-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(2.5858e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0003, 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(0.0002, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3669e-05, device='cuda:0', grad_fn=<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(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8369e-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(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3514e-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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5579e-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(0.0004, 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(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3804e-05, device='cuda:0', grad_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(1.6229e-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(4.1692e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5846e-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(0.0002, 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(2.1679e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-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(0.0001, 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(0.0014, 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(8.4044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-05, 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(3.5728e-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(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9537e-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(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-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(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2833e-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(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(2.1770e-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(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9875e-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(5.5710e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8135e-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(4.5931e-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(1.6229e-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(3.3094e-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(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(0.0003, 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(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2267e-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(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0684e-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(6.5007e-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(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(0.0005, 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(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-05, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8344e-05, device='cuda:0', grad_fn=<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(1.8561e-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(0.0009, 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(2.3799e-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(0.0004, 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(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(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5679e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1544e-05, 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(4.7505e-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(3.1005e-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(4.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-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(0.0001, 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(3.0157e-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(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6772e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, 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(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7979e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.9106e-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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1826e-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(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3899e-05, device='cuda:0', grad_fn=<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(1.7955e-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(0.0001, 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(3.1126e-05, 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(4.6416e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-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(0.0002, 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(0.0004, 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(6.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, 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(3.2125e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0421e-05, device='cuda:0', grad_fn=<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(2.1801e-05, 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(3.6546e-05, device='cuda:0', grad_fn=<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(1.8742e-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(0.0002, 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(7.0636e-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(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, 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(0.0001, 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(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1613e-05, device='cuda:0', grad_fn=<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(7.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-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(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-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(1.3837e-05, 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(3.4002e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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
				 {'precision': {'support': 0.7948717948717948, 'agreement': 0.5617977528089888, 'direct_attack': 0.36, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8532110091743119, 'agreement': 0.4716981132075472, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8230088495575222, 'agreement': 0.5128205128205128, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6727272727272727, 'recall': 0.6727272727272727, 'f1': 0.6727272727272727, 'support': None}, 'macro_avg': {'precision': 0.5191113236114431, 'recall': 0.47488210937665676, 'f1': 0.48710676118944046, 'support': None}, 'weighted_avg': {'precision': 0.6633085352132926, 'recall': 0.6727272727272727, 'f1': 0.6644511655187832, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2543e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2808e-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(2.3647e-05, device='cuda:0', grad_fn=<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(0.0012, 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(6.2884e-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(3.9115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(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(0.0004, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-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(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(5.1803e-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(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(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(9.2216e-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(0.0036, 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(7.8901e-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(6.2008e-05, 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(3.6667e-05, 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(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(0.0002, 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(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7467e-05, 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(8.3508e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8040e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6476e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<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(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-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(4.7869e-05, device='cuda:0', grad_fn=<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(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0584e-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(9.5886e-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(0.0005, 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(9.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-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(9.3688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8736e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8558e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<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(6.8052e-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(6.5065e-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(1.5351e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2976e-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(0.0003, 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(3.0490e-05, device='cuda:0', grad_fn=<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(3.5365e-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(0.0016, 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(0.0003, 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(0.0003, 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(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8404e-05, device='cuda:0', grad_fn=<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(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1813e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3036e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9761e-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(0.0016, 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(9.9339e-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(0.0008, 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(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(6.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0004, 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(1.4352e-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(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(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.2216e-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(0.0002, device='cuda:0', grad_fn=<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(0.0010, 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(0.0005, 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(2.8462e-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(0.0004, 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(3.7424e-05, 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(2.5810e-05, 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(9.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-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(1.4201e-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(6.1371e-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(6.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, 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(0.0001, 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(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-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(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(7.3417e-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(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5872e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3222e-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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8165e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-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(0.0004, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8135e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9271e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, 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(0.0004, 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(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7795e-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(1.8894e-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(5.3954e-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(6.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4701e-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(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(8.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-05, device='cuda:0', grad_fn=<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(7.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7246e-05, 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(1.7985e-05, 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(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6163e-05, device='cuda:0', grad_fn=<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(2.9552e-05, device='cuda:0', grad_fn=<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(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6955e-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(3.2549e-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(7.1724e-05, device='cuda:0', grad_fn=<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(9.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6023e-05, device='cuda:0', grad_fn=<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(0.0003, 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(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-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(0.0002, 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(0.0004, 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(0.0004, 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(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4326e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8855e-05, device='cuda:0', grad_fn=<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(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0007, 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(3.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3969e-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(3.5758e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-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(0.0002, device='cuda:0', grad_fn=<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(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-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(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-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(2.6312e-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(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1905e-05, 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(4.7869e-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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4059e-05, device='cuda:0', grad_fn=<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(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1784e-05, device='cuda:0', grad_fn=<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(1.4594e-05, 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(1.6714e-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(0.0002, 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(0.0001, 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(0.0008, 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(2.4011e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8414e-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(1.4201e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8514e-05, 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(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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2258e-05, 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(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4538e-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(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(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(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5828e-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(8.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8584e-05, 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(1.9923e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<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(0.0015, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-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(0.0004, 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(0.0001, 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(3.3760e-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(5.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-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(0.0006, 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(2.5706e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0004, 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(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9232e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5667e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-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(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(1.8621e-05, device='cuda:0', grad_fn=<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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6125e-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(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6006e-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(1.8591e-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(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(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(9.3946e-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(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-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(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7652e-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(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0001, 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(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<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(0.0002, 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(8.9096e-05, 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(8.8433e-05, 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(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(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8589e-05, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7164e-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(6.9703e-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(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-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(0.0010, 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(3.9452e-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(4.0696e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5901e-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(2.1316e-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(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(7.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0016e-05, 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(0.0011, 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(6.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5983e-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(0.0002, 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(6.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-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(0.0001, 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(7.4879e-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(0.0003, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(3.0157e-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(8.1623e-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(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6158e-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(5.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3118e-05, device='cuda:0', grad_fn=<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(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<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(0.0001, 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(8.4894e-05, 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(3.2549e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8654e-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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6796e-05, 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(0.0002, 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(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(3.3003e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9876e-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(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(0.0011, 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(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-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(1.7592e-05, device='cuda:0', grad_fn=<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(6.8493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(3.6394e-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(0.0006, 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(2.0892e-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(0.0005, 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(9.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5936e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, 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(0.0006, 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(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8614e-05, 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(4.8230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-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(0.0002, 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(1.0719e-05, 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(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7340e-05, device='cuda:0', grad_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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-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(4.9018e-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(5.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(9.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.6835e-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(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(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-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(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, 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(1.9166e-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(6.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0370e-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(6.8334e-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(9.1463e-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(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(2.7553e-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(0.0002, 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(0.0004, 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(6.8002e-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(4.0602e-05, device='cuda:0', grad_fn=<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(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4978e-05, device='cuda:0', grad_fn=<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(1.2868e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<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(2.7917e-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(3.1308e-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(1.0507e-05, 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(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(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3555e-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(3.6213e-05, device='cuda:0', grad_fn=<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(5.1258e-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(1.5260e-05, device='cuda:0', grad_fn=<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(2.8219e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(6.9968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<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(8.5153e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0002, 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(3.3094e-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(4.9624e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-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(5.0624e-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(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-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(2.4314e-05, device='cuda:0', grad_fn=<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(1.5049e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6791e-05, 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
				 {'precision': {'support': 0.7926136363636364, 'agreement': 0.5568181818181818, 'direct_attack': 0.36, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8532110091743119, 'agreement': 0.46226415094339623, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8217967599410898, 'agreement': 0.5051546391752577, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6709090909090909, 'recall': 0.6709090909090909, 'f1': 0.6709090909090909, 'support': None}, 'macro_avg': {'precision': 0.5176637777116501, 'recall': 0.4729953169238265, 'f1': 0.4853311685371029, 'support': None}, 'weighted_avg': {'precision': 0.6610062582002136, 'recall': 0.6709090909090909, 'f1': 0.6622531002261097, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3823e-05, device='cuda:0', grad_fn=<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(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-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(1.3504e-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(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2262e-05, 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(0.0005, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8890e-05, 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(5.9657e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0010, 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(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, 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(7.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2863e-05, device='cuda:0', grad_fn=<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(1.2899e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, 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(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0775e-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(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7264e-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(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-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(5.4931e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8844e-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(1.5018e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0516e-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(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7024e-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(0.0006, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-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(5.3137e-05, device='cuda:0', grad_fn=<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(1.9106e-05, device='cuda:0', grad_fn=<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(2.5948e-05, device='cuda:0', grad_fn=<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(9.6601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-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(3.7451e-05, device='cuda:0', grad_fn=<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(4.3084e-05, device='cuda:0', grad_fn=<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(0.0003, 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(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6804e-05, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6210e-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(3.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6717e-05, 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(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8043e-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(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2221e-05, device='cuda:0', grad_fn=<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(5.0428e-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(3.0793e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-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(4.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<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(8.2437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9390e-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(0.0016, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6347e-05, device='cuda:0', grad_fn=<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(3.2943e-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(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(0.0007, 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(2.2164e-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(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(6.2357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-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(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-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(0.0010, 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(9.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(6.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5456e-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(2.7765e-05, device='cuda:0', grad_fn=<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(1.3353e-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(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-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(2.2951e-05, device='cuda:0', grad_fn=<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(2.1225e-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(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9715e-05, device='cuda:0', grad_fn=<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(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4285e-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(0.0003, 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(0.0001, 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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(8.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-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(6.2647e-05, device='cuda:0', grad_fn=<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(6.3224e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1290e-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(7.6258e-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(1.3625e-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(2.2194e-05, 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(0.0001, 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(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-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(0.0006, 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(8.4047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8018e-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(5.6649e-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(6.0215e-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(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4056e-05, 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(6.6893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8718e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0370e-05, 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(9.0950e-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(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4698e-05, 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(7.2993e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4073e-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(0.0003, 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(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-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(0.0013, 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(6.5140e-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(0.0004, 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(0.0001, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7010e-05, 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(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(2.9341e-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(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0306e-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(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5310e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, 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(0.0002, 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(6.4700e-05, 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(2.9975e-05, device='cuda:0', grad_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(0.0007, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-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(7.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-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(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-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(2.1104e-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(0.0003, 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(2.2981e-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(4.4841e-05, device='cuda:0', grad_fn=<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(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, 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(3.1308e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4554e-05, 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(0.0002, 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(5.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(7.2329e-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(9.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8272e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5188e-05, 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(6.4690e-05, 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(1.5745e-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(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0968e-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(4.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0263e-05, device='cuda:0', grad_fn=<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(1.5382e-05, 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(0.0003, 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(8.6883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(4.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2612e-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(0.0001, 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(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-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(9.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-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(0.0004, 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(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, 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(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(1.3323e-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(9.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2218e-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(8.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-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(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-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(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3257e-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(9.0477e-05, 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(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6828e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6481e-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(9.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.2267e-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(6.1238e-05, device='cuda:0', grad_fn=<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(8.5992e-06, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2103e-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(1.5715e-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(7.8362e-05, device='cuda:0', grad_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(1.9560e-05, 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(5.5266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-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(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8847e-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(1.8107e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-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(0.0006, 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(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-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(0.0002, 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(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(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(7.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2440e-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(0.0002, device='cuda:0', grad_fn=<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(0.0007, 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(0.0001, 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(1.4292e-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(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-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(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-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(8.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-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(5.4256e-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(6.2781e-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(2.1316e-05, device='cuda:0', grad_fn=<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(7.1765e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9291e-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(8.5996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6530e-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(7.7322e-05, 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(7.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0505e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9984e-05, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2536e-05, device='cuda:0', grad_fn=<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(6.5243e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4992e-05, 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(6.3307e-05, device='cuda:0', grad_fn=<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(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-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(5.7862e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9824e-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(0.0002, 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(1.1960e-05, 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(1.5775e-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(1.4655e-05, device='cuda:0', grad_fn=<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(9.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-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(0.0001, 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(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-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(0.0001, 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(0.0004, 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(2.4223e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085e-05, 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(0.0001, 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(0.0002, 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(4.1754e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-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(1.6895e-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(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-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(0.0009, 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(0.0003, 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(0.0003, 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(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-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(6.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3723e-05, 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(2.4314e-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(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-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(0.0003, 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(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(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5037e-05, 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(0.0005, 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(1.2142e-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(0.0005, 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(0.0006, 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(9.8232e-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(6.7031e-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(4.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(6.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5627e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4998e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8215e-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(5.7799e-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(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, 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(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-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(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, 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(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5576e-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(0.0004, 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(0.0003, 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(3.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0017, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7959e-05, 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(2.8764e-05, 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(5.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8921e-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(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, 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(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(1.4292e-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
				 {'precision': {'support': 0.7932011331444759, 'agreement': 0.5568181818181818, 'direct_attack': 0.36, 'undercutter_attack': 0.34782608695652173, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.46226415094339623, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8235294117647058, 'agreement': 0.5051546391752577, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.31999999999999995, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6727272727272727, 'recall': 0.6727272727272727, 'f1': 0.6727272727272727, 'support': None}, 'macro_avg': {'precision': 0.5192613880761435, 'recall': 0.47360693771893353, 'f1': 0.48631136226816274, 'support': None}, 'weighted_avg': {'precision': 0.6620821516903637, 'recall': 0.6727272727272727, 'f1': 0.6635943115992611, 'support': None}}
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9310e-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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1622e-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(1.4867e-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>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-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(8.6437e-05, device='cuda:0', grad_fn=<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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, 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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7640e-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(9.0219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-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(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8051e-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(1.4110e-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(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-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(1.9590e-05, device='cuda:0', grad_fn=<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(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8303e-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(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-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(1.8107e-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(0.0005, 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(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, 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(0.0004, 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(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(4.0527e-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(3.1459e-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(0.0004, 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(7.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, 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(2.8734e-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(5.1319e-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(6.3705e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3570e-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(0.0001, 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(9.1636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1207e-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(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(1.9590e-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(3.3185e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-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(0.0005, 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(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-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(9.6540e-05, device='cuda:0', grad_fn=<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(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9943e-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(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-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(0.0012, 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(1.6078e-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(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6807e-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(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0249e-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(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6077e-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(0.0002, 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(0.0002, 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(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, 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(2.7553e-05, device='cuda:0', grad_fn=<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(5.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2944e-05, device='cuda:0', grad_fn=<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(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-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(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-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(0.0002, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-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(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(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-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(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(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5147e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1128e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3747e-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(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2253e-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(0.0002, device='cuda:0', grad_fn=<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(7.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7966e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9393e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-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(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0104e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6289e-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(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-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(1.2021e-05, 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(1.0416e-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(9.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-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(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, 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(0.0002, 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(5.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1178e-05, 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(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9766e-05, 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(0.0004, 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(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9545e-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(8.5158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1356e-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(5.9887e-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(3.2256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3881e-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(0.0001, 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(9.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0326e-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(3.7135e-05, 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(0.0004, 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(1.1657e-05, device='cuda:0', grad_fn=<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(1.9499e-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(1.9832e-05, 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(5.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7401e-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(2.3526e-05, device='cuda:0', grad_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(0.0004, 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(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, 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(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2600e-05, 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(0.0001, 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(2.2890e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-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(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(4.6989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3437e-05, device='cuda:0', grad_fn=<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(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8282e-05, 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(3.2536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6929e-05, 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(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5933e-05, device='cuda:0', grad_fn=<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(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8472e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-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(2.1588e-05, device='cuda:0', grad_fn=<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(2.5282e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2189e-05, device='cuda:0', grad_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(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-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(8.8030e-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(7.5446e-05, device='cuda:0', grad_fn=<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(1.4504e-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(9.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8736e-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(9.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8083e-05, device='cuda:0', grad_fn=<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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-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(0.0005, 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(1.5200e-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(3.9209e-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(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(6.5034e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4879e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-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(0.0006, 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(0.0004, 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(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7611e-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(1.2354e-05, 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(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8665e-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(9.2863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5998e-05, device='cuda:0', grad_fn=<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(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6204e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9071e-05, 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(6.2940e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7104e-05, device='cuda:0', grad_fn=<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(6.9181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0574e-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(8.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0377e-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(1.2051e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8281e-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(0.0006, 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(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-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(1.6714e-05, device='cuda:0', grad_fn=<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(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3556e-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(6.5880e-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(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(1.7259e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5566e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-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(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4920e-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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7982e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0441e-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(9.6287e-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(0.0001, 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(0.0001, 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(4.5839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3438e-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(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-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(1.6108e-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(3.9694e-05, device='cuda:0', grad_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(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6592e-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(0.0002, 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(4.6536e-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(5.5467e-05, device='cuda:0', grad_fn=<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(2.4495e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6803e-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(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9075e-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(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, 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(9.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6869e-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(9.2786e-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(1.3656e-05, 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(1.8016e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4727e-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(0.0001, 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(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8165e-05, 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(4.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<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(2.2951e-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(4.9438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4504e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(1.0961e-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(0.0004, 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(0.0001, 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(8.4315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7999e-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(0.0006, 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(1.7652e-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(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-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(7.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, 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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-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(7.1787e-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(9.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7229e-05, device='cuda:0', grad_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(0.0006, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7697e-05, 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(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4751e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4783e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3111e-05, device='cuda:0', grad_fn=<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(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-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(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-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(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9713e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(5.4799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-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(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8067e-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(2.3981e-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(1.1143e-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(7.9538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3920e-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(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(0.0006, 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(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0583e-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(1.5291e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-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(6.9395e-05, device='cuda:0', grad_fn=<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(5.3923e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4659e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5022e-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(1.1839e-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(1.6229e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-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(3.3457e-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(4.7838e-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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-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(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(6.3610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1611e-05, device='cuda:0', grad_fn=<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(5.1943e-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
				 {'precision': {'support': 0.7932011331444759, 'agreement': 0.5568181818181818, 'direct_attack': 0.3673469387755102, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.46226415094339623, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8235294117647058, 'agreement': 0.5051546391752577, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6727272727272727, 'recall': 0.6727272727272727, 'f1': 0.6727272727272727, 'support': None}, 'macro_avg': {'precision': 0.51925066482292, 'recall': 0.47360693771893353, 'f1': 0.4865970043155136, 'support': None}, 'weighted_avg': {'precision': 0.6618765164903582, 'recall': 0.6727272727272727, 'f1': 0.6636091760479124, 'support': None}}
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<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(9.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<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(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5120e-05, device='cuda:0', grad_fn=<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(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4262e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9259e-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(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-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(9.5344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-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(0.0001, 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(0.0003, 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(7.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6145e-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(6.1246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(8.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9040e-05, device='cuda:0', grad_fn=<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(1.5836e-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(1.3080e-05, 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(8.5112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2802e-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(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-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(1.8107e-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(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8948e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, 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(9.8406e-06, 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(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7481e-05, 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(0.0001, 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(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3060e-05, device='cuda:0', grad_fn=<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(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8581e-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(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, 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(0.0003, 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(1.7501e-05, 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(0.0005, 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(0.0003, 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(9.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9835e-05, device='cuda:0', grad_fn=<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(8.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-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(1.5200e-05, 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(0.0002, 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(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7944e-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(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2352e-05, 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(2.6433e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.8409e-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(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4200e-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(5.2580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-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(1.1203e-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(0.0007, 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(1.1627e-05, device='cuda:0', grad_fn=<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(2.3859e-05, 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(2.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0721e-05, 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(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-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(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<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(3.4910e-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(4.2296e-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(9.0255e-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(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(0.0002, 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(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-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(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(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-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(7.1655e-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(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4543e-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(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3376e-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(2.0741e-05, 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(0.0003, 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(4.9715e-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(1.7107e-05, device='cuda:0', grad_fn=<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(1.3444e-05, 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(0.0001, 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(9.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3783e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(8.7625e-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(0.0002, 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(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9599e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, 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(0.0003, 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(8.5400e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7627e-05, 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(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(2.8673e-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(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(0.0005, 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(0.0002, 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(5.0490e-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(0.0004, 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(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8339e-05, 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(7.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6135e-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(1.0113e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, 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(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(5.7428e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7319e-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(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5869e-05, device='cuda:0', grad_fn=<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(6.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6160e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3112e-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(5.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-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(6.4716e-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(8.0542e-06, 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(3.2185e-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(8.5138e-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(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-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(2.5161e-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(8.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-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(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-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(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4130e-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(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3650e-05, 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(9.0903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8497e-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(9.0686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5641e-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(1.2293e-05, device='cuda:0', grad_fn=<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(3.1305e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, 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(3.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2383e-05, 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(6.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4040e-05, device='cuda:0', grad_fn=<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(8.5150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5620e-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(4.4126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5095e-05, device='cuda:0', grad_fn=<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(9.3605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7126e-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(2.4344e-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(0.0002, 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(2.8401e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5194e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, 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(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1367e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-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(0.0001, 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(5.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-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(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8737e-05, device='cuda:0', grad_fn=<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(2.4556e-05, 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(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9411e-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(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0839e-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(0.0007, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2723e-05, 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(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-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(9.3536e-05, device='cuda:0', grad_fn=<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(0.0002, 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(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(1.2384e-05, device='cuda:0', grad_fn=<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(1.4988e-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(0.0003, 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(3.8132e-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(0.0002, 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(1.6078e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-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(8.5377e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2649e-05, 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(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-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(7.6021e-05, device='cuda:0', grad_fn=<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(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9286e-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(1.6169e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5422e-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(3.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-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(1.2899e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7595e-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(0.0001, 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(0.0001, 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(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9926e-05, device='cuda:0', grad_fn=<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(5.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<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(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5157e-05, device='cuda:0', grad_fn=<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(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1903e-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(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8995e-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(8.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4386e-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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, 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(0.0001, 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(0.0002, 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(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0159e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, 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(9.2542e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1505e-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(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6874e-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(1.4382e-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(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9556e-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(8.3873e-06, 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(1.2172e-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(1.5745e-05, 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(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6838e-05, 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(4.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5142e-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(6.4639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5556e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-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(3.6740e-05, 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(0.0007, 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(5.2116e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0529e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, 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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-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(1.6229e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7308e-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(3.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0470e-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(8.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5967e-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(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8464e-05, device='cuda:0', grad_fn=<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(0.0004, 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(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2893e-05, device='cuda:0', grad_fn=<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(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1866e-05, 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(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4581e-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(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9018e-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(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9282e-05, 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(1.1688e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-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(9.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(9.0728e-05, device='cuda:0', grad_fn=<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(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6173e-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(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1690e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7002e-05, 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(6.2584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-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(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-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(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7113e-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(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(1.1173e-05, device='cuda:0', grad_fn=<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(9.5182e-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(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(9.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8772e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9312e-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(7.4795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-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(9.0988e-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
				 {'precision': {'support': 0.7937853107344632, 'agreement': 0.5632183908045977, 'direct_attack': 0.3673469387755102, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8593272171253823, 'agreement': 0.46226415094339623, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8252569750367107, 'agreement': 0.5077720207253887, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6745454545454546, 'recall': 0.6745454545454546, 'f1': 0.6745454545454546, 'support': None}, 'macro_avg': {'precision': 0.5206475421382006, 'recall': 0.47421855851404054, 'f1': 0.4874659932799408, 'support': None}, 'weighted_avg': {'precision': 0.6634573314439688, 'recall': 0.6745454545454546, 'f1': 0.665140731746566, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-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(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-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(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(1.6744e-05, device='cuda:0', grad_fn=<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(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-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(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(6.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-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(8.0542e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, 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(8.8931e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4631e-05, 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(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1586e-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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1266e-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(2.0578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4134e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-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(0.0001, 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(3.9203e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1228e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-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(3.4424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0454e-05, device='cuda:0', grad_fn=<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(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3872e-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(1.0840e-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>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1888e-05, device='cuda:0', grad_fn=<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(1.5654e-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(0.0005, 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(7.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3782e-05, 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(3.1367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2416e-05, 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(9.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6268e-05, device='cuda:0', grad_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(4.6081e-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(0.0010, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, 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(7.0720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8816e-05, device='cuda:0', grad_fn=<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(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0404e-05, 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(3.2120e-05, device='cuda:0', grad_fn=<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(3.0520e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7758e-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(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, 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(0.0001, 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(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1307e-05, device='cuda:0', grad_fn=<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(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-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(8.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9261e-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(8.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9980e-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(8.9200e-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(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1255e-05, device='cuda:0', grad_fn=<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(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, 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(8.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8038e-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(0.0002, 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(8.2713e-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(2.9218e-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(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2998e-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(9.1442e-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(0.0001, 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(6.8244e-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(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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2849e-05, device='cuda:0', grad_fn=<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(8.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-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(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<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(7.8564e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-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(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7133e-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(2.6584e-05, device='cuda:0', grad_fn=<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(1.9136e-05, device='cuda:0', grad_fn=<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(0.0001, 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(8.9962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6717e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-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(3.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-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(4.2334e-05, device='cuda:0', grad_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(0.0002, 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(1.5412e-05, 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(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(8.8844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4198e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-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(8.8178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, 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(4.1269e-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(0.0003, 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(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3087e-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(9.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5607e-05, 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(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(9.7929e-05, device='cuda:0', grad_fn=<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(4.2902e-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(2.0408e-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(6.3970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4974e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4508e-05, 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(0.0003, 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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3270e-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(2.3011e-05, device='cuda:0', grad_fn=<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(7.4830e-05, 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(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-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(9.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(6.3307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0885e-05, 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(0.0003, 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(1.5563e-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(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7008e-05, 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(6.5371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3206e-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(5.3472e-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(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3537e-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(7.8413e-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(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3416e-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(3.0223e-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(0.0003, 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(4.9906e-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(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0005, 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(0.0003, 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(1.0446e-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(8.4478e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<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(6.2790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7216e-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(8.2219e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7202e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4317e-05, device='cuda:0', grad_fn=<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(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3481e-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(7.6000e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5384e-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(1.4443e-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(8.2928e-05, device='cuda:0', grad_fn=<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(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5012e-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(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4704e-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(8.0123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-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(8.0684e-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(2.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9941e-05, device='cuda:0', grad_fn=<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(3.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1999e-05, device='cuda:0', grad_fn=<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(9.8390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, 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(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2829e-05, device='cuda:0', grad_fn=<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(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-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(5.0827e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1959e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, 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(1.0840e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1092e-05, device='cuda:0', grad_fn=<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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2069e-05, device='cuda:0', grad_fn=<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(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4655e-05, 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(2.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0060e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9490e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2589e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1602e-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(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(7.3674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8753e-05, device='cuda:0', grad_fn=<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(6.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7359e-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(9.3755e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6747e-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(0.0005, 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(9.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4230e-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(4.1545e-05, device='cuda:0', grad_fn=<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(6.1590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1525e-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(0.0002, 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(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-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(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1864e-05, device='cuda:0', grad_fn=<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(8.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3937e-05, device='cuda:0', grad_fn=<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(7.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(9.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5332e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7904e-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(0.0003, 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(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(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2784e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7793e-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(1.2263e-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(3.2196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6732e-05, device='cuda:0', grad_fn=<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(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2945e-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(9.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2801e-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(8.1029e-05, 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(5.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(2.6736e-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(9.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4516e-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(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6421e-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(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8284e-05, device='cuda:0', grad_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(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3217e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-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(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9381e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1641e-05, device='cuda:0', grad_fn=<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(6.6033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8107e-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(0.0017, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-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(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-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(2.0014e-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(4.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, 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(7.8467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4681e-05, device='cuda:0', grad_fn=<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(3.4956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-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(5.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2957e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8799e-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(1.3050e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8017e-05, 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(4.0329e-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(1.1567e-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(0.0005, 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(5.9680e-05, device='cuda:0', grad_fn=<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(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3511e-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(3.7726e-05, 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(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7523e-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(0.0001, 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(8.7712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2950e-05, 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(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9637e-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(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(0.0002, 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(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, 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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5285e-05, device='cuda:0', grad_fn=<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(1.9136e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-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(5.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1133e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4933e-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(3.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0832e-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(1.6381e-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(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-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(0.0001, 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>)
			------------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
				 {'precision': {'support': 0.7915492957746478, 'agreement': 0.5581395348837209, 'direct_attack': 0.3673469387755102, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8593272171253823, 'agreement': 0.4528301886792453, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8240469208211143, 'agreement': 0.5, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6727272727272727, 'recall': 0.6727272727272727, 'f1': 0.6727272727272727, 'support': None}, 'macro_avg': {'precision': 0.5191845679620621, 'recall': 0.4723317660612104, 'f1': 0.4856695782917438, 'support': None}, 'weighted_avg': {'precision': 0.6611490848631097, 'recall': 0.6727272727272727, 'f1': 0.6629234191549456, 'support': None}}
Loss: tensor(9.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-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(8.0699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9973e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-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(7.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0083e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6378e-05, device='cuda:0', grad_fn=<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(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5216e-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(2.7432e-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(4.3960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5306e-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(0.0002, 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(6.1643e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6600e-05, 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(0.0001, 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(6.3885e-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(2.1031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-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(8.1994e-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(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6501e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, 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(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(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8462e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9683e-05, device='cuda:0', grad_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(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-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(1.1233e-05, 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(8.1803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0085e-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(5.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7520e-05, 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(1.5472e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9294e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<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(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0324e-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(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(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5222e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-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(5.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3799e-05, 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(6.5019e-05, 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(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6607e-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(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2187e-05, 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(6.2084e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8392e-05, device='cuda:0', grad_fn=<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(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, 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(1.6895e-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(8.2359e-06, 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(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1235e-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(8.5241e-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(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-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(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8875e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8878e-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(6.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9057e-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(7.2313e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6735e-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(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1875e-05, device='cuda:0', grad_fn=<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(5.2803e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8058e-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(2.0922e-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(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, 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(7.1711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5547e-05, device='cuda:0', grad_fn=<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(1.8349e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-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(7.8230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4900e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-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(1.6290e-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(2.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5289e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9281e-05, device='cuda:0', grad_fn=<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(9.0870e-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(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-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(1.2929e-05, 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(9.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2010e-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(1.9045e-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(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-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(6.7031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3843e-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(8.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3821e-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(1.2414e-05, device='cuda:0', grad_fn=<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(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0987e-05, 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(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7703e-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(3.8381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6066e-05, 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(1.0900e-05, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5058e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4828e-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(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4667e-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(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-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(7.0541e-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(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(6.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6243e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4193e-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(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5498e-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(4.4309e-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(6.2182e-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(9.3110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215e-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(6.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2908e-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(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-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(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6121e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-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(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(5.6616e-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(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, 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(5.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2898e-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(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(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(1.4110e-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(0.0002, 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(8.4363e-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(1.0325e-05, 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(5.5015e-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(9.3179e-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(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, 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(8.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2374e-05, 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(1.2566e-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(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7477e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, 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(5.4619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9612e-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(0.0004, 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(0.0001, 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(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2424e-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(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(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4807e-05, device='cuda:0', grad_fn=<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(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1340e-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(0.0003, 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(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4864e-05, 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(8.8398e-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(7.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4972e-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(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(4.6964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, 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(9.5557e-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(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9604e-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(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2928e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-05, device='cuda:0', grad_fn=<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(6.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8461e-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(4.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-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(9.2046e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5014e-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(0.0003, 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(2.6923e-05, device='cuda:0', grad_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(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7106e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-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(0.0003, 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(8.5280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1123e-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(6.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9541e-05, device='cuda:0', grad_fn=<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(8.3405e-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(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0071e-05, 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(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4269e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-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(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0468e-05, 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(9.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-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(0.0002, 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(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5786e-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(2.4056e-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(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(2.1184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0523e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5984e-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(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6529e-05, device='cuda:0', grad_fn=<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(1.8197e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-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(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(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-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(6.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5863e-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(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(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2808e-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(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8692e-05, device='cuda:0', grad_fn=<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(9.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9109e-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(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8537e-05, 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(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(9.6618e-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(2.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0973e-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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-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(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5046e-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(8.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2662e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1976e-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(5.9158e-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(1.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5365e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-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(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5824e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9264e-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(6.9036e-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(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, 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(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(9.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0865e-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(4.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1531e-05, device='cuda:0', grad_fn=<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(0.0004, 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(7.5092e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7280e-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(6.6184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9763e-05, 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(3.5837e-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(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-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(8.7762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3820e-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(1.1415e-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(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9636e-05, 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(3.0369e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9218e-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(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-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
				 {'precision': {'support': 0.7893258426966292, 'agreement': 0.5529411764705883, 'direct_attack': 0.3673469387755102, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8593272171253823, 'agreement': 0.44339622641509435, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8228404099560762, 'agreement': 0.49214659685863876, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6709090909090909, 'recall': 0.6709090909090909, 'f1': 0.6709090909090909, 'support': None}, 'macro_avg': {'precision': 0.5177002056638319, 'recall': 0.4704449736083802, 'f1': 0.48385759549046387, 'support': None}, 'weighted_avg': {'precision': 0.6588252755025567, 'recall': 0.6709090909090909, 'f1': 0.6606925286352152, 'support': None}}
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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8437e-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(5.9307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9855e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7340e-05, 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(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(8.3570e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, 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(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, 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(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5005e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8489e-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(0.0004, 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(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6416e-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(1.6772e-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(4.0106e-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(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0156e-05, device='cuda:0', grad_fn=<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(3.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, 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(7.2147e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-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(4.4386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5922e-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(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(7.4696e-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(8.9087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0956e-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(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4132e-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(4.1785e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8563e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-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(7.0238e-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(0.0001, 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(8.3795e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4342e-05, device='cuda:0', grad_fn=<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(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-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(3.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, 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(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4464e-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(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2031e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2517e-05, 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(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(5.1654e-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(4.6255e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8559e-05, device='cuda:0', grad_fn=<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(7.9331e-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(5.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4004e-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(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, 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(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-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(1.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7165e-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(7.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8906e-05, device='cuda:0', grad_fn=<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(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-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(9.8105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-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(7.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3522e-05, 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(0.0001, 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(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9204e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, 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(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(6.8432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1571e-05, 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(1.3262e-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(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-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(1.0234e-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(1.0810e-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(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6908e-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(2.7553e-05, device='cuda:0', grad_fn=<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(1.0325e-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(7.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6983e-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(8.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, 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(1.0355e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-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(1.5260e-05, device='cuda:0', grad_fn=<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(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1038e-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(8.9625e-06, 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(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5928e-05, device='cuda:0', grad_fn=<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(3.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-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(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-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(8.2359e-06, 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(1.0870e-05, device='cuda:0', grad_fn=<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(8.0296e-05, 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(9.5216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, 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(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-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(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(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-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(9.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8866e-05, 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(2.8916e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3796e-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(0.0002, 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(4.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, 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(9.3645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6878e-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(1.3474e-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(7.2367e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, 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(0.0001, 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(5.5702e-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(0.0002, 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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3865e-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(5.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4722e-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(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0546e-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(9.4708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9845e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8851e-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(1.9015e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8143e-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(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<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(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, 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(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, 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(7.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1426e-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(3.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6893e-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(4.4769e-05, 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(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3592e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5660e-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(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5056e-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(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7943e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3981e-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(6.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0261e-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(0.0002, 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(5.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-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(3.7522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0204e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2916e-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(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0602e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9667e-05, device='cuda:0', grad_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(9.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5889e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9689e-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(0.0001, 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(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3064e-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(7.0550e-06, 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(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, 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(2.9689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-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(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7618e-05, 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(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9675e-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(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8182e-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(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-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(0.0002, 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(1.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3653e-05, device='cuda:0', grad_fn=<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(5.4879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4090e-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(5.7530e-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(5.4007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3841e-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(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, 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(5.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2049e-05, 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(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(4.3842e-05, 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(2.1831e-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(1.1021e-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(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7188e-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(9.9315e-06, 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(9.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0849e-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(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4172e-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(1.1930e-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(8.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4335e-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(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9745e-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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8251e-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(9.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, 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(5.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7415e-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(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1422e-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(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6947e-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(2.5965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6423e-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(9.9357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2464e-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(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5971e-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(8.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-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(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1955e-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(6.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4764e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, 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(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6991e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-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(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7712e-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(3.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2173e-05, 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(2.4344e-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(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-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(1.3323e-05, 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(8.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-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(3.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8261e-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(7.2673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7868e-05, device='cuda:0', grad_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(8.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7951e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8257e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5526e-05, device='cuda:0', grad_fn=<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(2.7210e-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(7.9331e-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(2.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-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(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3511e-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>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2458e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5637e-05, 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(4.8565e-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(7.4962e-05, device='cuda:0', grad_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-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4068e-05, device='cuda:0', grad_fn=<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(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1873e-05, 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(8.6293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7693e-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(4.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, 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(4.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0755e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, 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(1.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-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(1.6865e-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(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7153e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3819e-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(6.6008e-06, 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(9.0071e-05, 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>)
			------------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
				 {'precision': {'support': 0.7893258426966292, 'agreement': 0.5529411764705883, 'direct_attack': 0.3673469387755102, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8593272171253823, 'agreement': 0.44339622641509435, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8228404099560762, 'agreement': 0.49214659685863876, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6709090909090909, 'recall': 0.6709090909090909, 'f1': 0.6709090909090909, 'support': None}, 'macro_avg': {'precision': 0.5177002056638319, 'recall': 0.4704449736083802, 'f1': 0.48385759549046387, 'support': None}, 'weighted_avg': {'precision': 0.6588252755025567, 'recall': 0.6709090909090909, 'f1': 0.6606925286352152, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6592e-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(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(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-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(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4228e-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(3.8119e-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(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(6.0861e-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(9.0740e-05, device='cuda:0', grad_fn=<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(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0150e-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(8.7885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, 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(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3332e-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(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, 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(6.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0435e-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(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1631e-05, device='cuda:0', grad_fn=<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(2.9431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7381e-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(4.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9250e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6518e-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(7.5395e-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(2.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8291e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0836e-05, device='cuda:0', grad_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(0.0001, 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(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4440e-05, 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(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-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(0.0003, 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(5.2018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1304e-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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5192e-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(3.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-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(8.2256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7025e-05, device='cuda:0', grad_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(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2299e-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(3.0974e-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(4.1634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-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(9.0534e-06, 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(4.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4656e-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(9.1442e-06, 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(0.0002, 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(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2476e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, 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(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-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(7.7623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-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(7.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6836e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5966e-05, 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(7.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-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(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4227e-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(1.0931e-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(0.0001, 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(8.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1088e-05, device='cuda:0', grad_fn=<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(5.6829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5218e-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(4.6173e-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(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9277e-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(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2274e-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(1.7652e-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(4.3750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, 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(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-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(7.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(8.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(8.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2808e-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(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6606e-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(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, 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(7.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-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(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0588e-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(2.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-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(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2600e-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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3184e-05, 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(0.0002, 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(1.7350e-05, device='cuda:0', grad_fn=<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(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-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(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5523e-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(8.9939e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5842e-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(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9762e-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(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-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(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7119e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1490e-05, 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(4.4718e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7794e-05, 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(2.2931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1040e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4866e-05, device='cuda:0', grad_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(3.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0208e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-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(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7950e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1142e-05, device='cuda:0', grad_fn=<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(7.4486e-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(7.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0604e-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(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, 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(5.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9406e-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(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5880e-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(6.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4837e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7482e-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(9.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-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(8.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-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(5.3765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2012e-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(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9397e-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(8.0673e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1428e-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(2.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-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(4.0747e-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(8.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1730e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9255e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-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(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, 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(1.1445e-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(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, 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(0.0001, 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(6.4494e-06, 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(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4977e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9031e-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(3.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0724e-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(5.2715e-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(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8523e-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(3.8079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-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(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_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(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(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(1.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7923e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3152e-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(7.8179e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-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(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(3.8572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6681e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7798e-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(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5408e-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(3.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1618e-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(7.6303e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-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(8.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-05, 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(8.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4757e-05, device='cuda:0', grad_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(1.1143e-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(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3397e-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(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(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-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(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8498e-05, 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(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9056e-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(8.8676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1400e-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(6.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, 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(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2251e-05, device='cuda:0', grad_fn=<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(4.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3630e-05, device='cuda:0', grad_fn=<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(9.8406e-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(3.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(7.3675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3839e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9216e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9918e-05, device='cuda:0', grad_fn=<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(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6677e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3137e-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(6.3283e-06, 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(6.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5123e-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(2.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2540e-05, 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(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0068e-05, 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(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0489e-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(1.3777e-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(9.8239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4637e-05, 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(9.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3763e-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(5.0228e-05, device='cuda:0', grad_fn=<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(3.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8908e-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(6.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-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(6.3553e-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(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7051e-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(3.8367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8244e-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(6.0215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-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
				 {'precision': {'support': 0.7887323943661971, 'agreement': 0.5529411764705883, 'direct_attack': 0.36, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.44339622641509435, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8211143695014662, 'agreement': 0.49214659685863876, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6690909090909091, 'recall': 0.6690909090909091, 'f1': 0.6690909090909091, 'support': None}, 'macro_avg': {'precision': 0.5161121282426435, 'recall': 0.4698333528132732, 'f1': 0.48259308198585443, 'support': None}, 'weighted_avg': {'precision': 0.6579514787456545, 'recall': 0.6690909090909091, 'f1': 0.6593403835728033, 'support': None}}
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5618e-05, 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(7.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7824e-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(1.0688e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-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(7.1967e-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(1.6078e-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(4.0059e-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(8.0845e-06, 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(4.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8361e-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(7.4606e-05, 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(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3515e-05, device='cuda:0', grad_fn=<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(2.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5679e-05, 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(8.0202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8974e-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(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1844e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, 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(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4004e-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(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1844e-05, device='cuda:0', grad_fn=<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(4.6811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0357e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6441e-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(8.2793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9026e-05, device='cuda:0', grad_fn=<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(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4682e-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(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3970e-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(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-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(6.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7639e-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(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6054e-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(3.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-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(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6233e-05, 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(2.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2540e-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(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-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(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6215e-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(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, 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(8.9250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, 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(2.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0213e-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(3.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0918e-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(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<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(9.0231e-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(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0151e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0482e-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(6.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, 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(6.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, 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(6.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1237e-05, device='cuda:0', grad_fn=<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(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6386e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, 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(1.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9696e-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(9.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-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(4.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4700e-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(4.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, 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(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6777e-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(7.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4969e-05, device='cuda:0', grad_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(0.0002, 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(0.0002, 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(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-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(3.7546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-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(6.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7982e-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(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, 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(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8109e-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(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8327e-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(6.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4013e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9041e-05, device='cuda:0', grad_fn=<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(7.9324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5535e-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(6.1554e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6167e-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(9.9012e-06, 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(7.2832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, 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(0.0002, 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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8512e-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(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, 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(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, 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(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-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(8.3446e-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(1.3570e-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(1.5472e-05, 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(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0129e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0748e-05, 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(7.8873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3372e-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(0.0002, device='cuda:0', grad_fn=<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(8.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4490e-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(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0466e-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(7.8120e-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(8.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4273e-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(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3870e-05, 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(7.6703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2556e-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(3.7206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7161e-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(5.7530e-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(3.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-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(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7390e-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(6.4960e-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(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(4.7078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3259e-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(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, 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(8.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5267e-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(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1938e-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(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, 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(9.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4978e-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(7.4486e-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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6377e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, 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(5.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8961e-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(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-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(5.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-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(1.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6491e-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(6.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0644e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3395e-05, 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(5.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2897e-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(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7795e-05, 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(3.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7899e-05, 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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9835e-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(8.7591e-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(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9301e-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(4.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-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(1.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-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(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3941e-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(6.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, 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(1.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9062e-05, device='cuda:0', grad_fn=<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(9.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4929e-05, 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(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1294e-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(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4994e-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(9.4576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5039e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3182e-05, 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(6.2593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7356e-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(5.9044e-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(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3759e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8321e-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(7.1761e-06, 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(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3805e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2801e-05, 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(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4417e-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(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8780e-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(4.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-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(9.0349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2665e-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(4.3208e-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(8.5997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9067e-05, 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(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7640e-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(4.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1050e-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(8.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5802e-05, device='cuda:0', grad_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(7.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6639e-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(4.8277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7720e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-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(2.0134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8391e-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
				 {'precision': {'support': 0.7887323943661971, 'agreement': 0.5529411764705883, 'direct_attack': 0.36, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.44339622641509435, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8211143695014662, 'agreement': 0.49214659685863876, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6690909090909091, 'recall': 0.6690909090909091, 'f1': 0.6690909090909091, 'support': None}, 'macro_avg': {'precision': 0.5161121282426435, 'recall': 0.4698333528132732, 'f1': 0.48259308198585443, 'support': None}, 'weighted_avg': {'precision': 0.6579514787456545, 'recall': 0.6690909090909091, 'f1': 0.6593403835728033, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2233e-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(8.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5394e-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(7.8008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6210e-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(0.0002, 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(7.0322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9706e-05, device='cuda:0', grad_fn=<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(4.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0428e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7016e-05, device='cuda:0', grad_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(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9358e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-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(1.6562e-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(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2294e-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(0.0001, 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(1.3264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, 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(5.1163e-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(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3698e-05, 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(9.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5557e-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(4.5439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6119e-06, 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(5.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-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(7.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2530e-05, device='cuda:0', grad_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(5.5391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4182e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2273e-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(0.0001, 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(9.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-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(7.9700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6570e-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(3.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-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(6.1379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2246e-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(7.6908e-06, 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(5.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-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(9.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3826e-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(8.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-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(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1366e-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(0.0001, 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(3.8275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4782e-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(7.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8427e-05, device='cuda:0', grad_fn=<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(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6252e-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(4.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6117e-05, 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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0298e-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(6.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, 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(3.8861e-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(1.5142e-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(4.4362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-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(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, 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(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, 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(1.7411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871e-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(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3564e-05, device='cuda:0', grad_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(3.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0870e-05, 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(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2265e-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(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9658e-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(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3742e-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(4.3739e-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(5.9437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8252e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9873e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4327e-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(3.3737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9833e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, 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(8.7641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-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(1.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3647e-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(2.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-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(6.8351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-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(3.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6978e-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(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7400e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6099e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2596e-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(5.6114e-05, 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(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5325e-05, 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(5.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2550e-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(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2140e-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(5.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, 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(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0278e-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(1.2172e-05, 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(5.6936e-05, device='cuda:0', grad_fn=<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(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6686e-05, 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(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3833e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, 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(6.9501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9491e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3881e-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(0.0001, 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(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4666e-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(8.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, 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(9.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, 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(3.6647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6528e-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(2.8068e-05, 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(7.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1584e-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(1.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, 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(2.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6324e-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(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, 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(4.9658e-06, 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(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-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(5.0263e-06, 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(8.4756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8705e-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(2.5724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6868e-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(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-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(3.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4411e-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(3.0197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0882e-05, 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(8.9682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3912e-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(2.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-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(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7881e-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(4.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, 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(5.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8850e-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(9.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4074e-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(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, 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(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(3.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6150e-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(0.0002, 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(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, 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(3.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5821e-05, 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(6.0173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, 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(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2516e-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(5.5127e-05, device='cuda:0', grad_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(5.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-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(6.3464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6075e-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
				 {'precision': {'support': 0.7887323943661971, 'agreement': 0.5529411764705883, 'direct_attack': 0.36, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.44339622641509435, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8211143695014662, 'agreement': 0.49214659685863876, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6690909090909091, 'recall': 0.6690909090909091, 'f1': 0.6690909090909091, 'support': None}, 'macro_avg': {'precision': 0.5161121282426435, 'recall': 0.4698333528132732, 'f1': 0.48259308198585443, 'support': None}, 'weighted_avg': {'precision': 0.6579514787456545, 'recall': 0.6690909090909091, 'f1': 0.6593403835728033, 'support': None}}
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-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(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4843e-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(4.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3389e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9453e-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(3.5989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4574e-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(8.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9236e-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(1.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4116e-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(1.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-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(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9948e-05, 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(6.2045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2060e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6258e-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(8.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-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(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2904e-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(6.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0994e-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(3.0821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9306e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5822e-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(5.1996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5526e-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(6.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2381e-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(2.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4614e-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(7.6908e-06, 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(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1939e-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(9.3960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7785e-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(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2016e-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(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7353e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0561e-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(4.7360e-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(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, 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(2.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3494e-05, 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(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-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(4.3566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8119e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9764e-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(4.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-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(2.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0152e-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(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-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(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8443e-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(4.2867e-05, 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(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3209e-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(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, 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(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0747e-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(9.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7492e-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(7.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2040e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9455e-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(9.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9712e-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(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9068e-05, 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(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2429e-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(1.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0634e-05, 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(1.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2638e-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(3.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2772e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3955e-05, 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(5.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0389e-05, 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(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4456e-05, 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(2.0543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0301e-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(3.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7087e-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(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1623e-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(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-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(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7285e-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(1.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8106e-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(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5428e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, 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(2.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8767e-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(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-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(3.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-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(5.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1374e-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(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7945e-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(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4930e-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(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, 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(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-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(8.9040e-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(4.7870e-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(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, 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(4.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2149e-05, 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(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-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(2.1700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-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(5.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7576e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-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(1.2299e-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(9.3674e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, 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(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5942e-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(1.0880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4683e-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(1.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, 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(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3527e-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(6.5765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1725e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9157e-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(2.2620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2956e-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(2.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8784e-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(4.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1178e-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(2.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0989e-05, 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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5970e-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(4.9355e-06, 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(2.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9036e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8222e-05, 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(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2582e-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(7.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7205e-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(5.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5231e-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(5.3594e-06, 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(9.2500e-05, 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(3.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3901e-05, 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(8.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7942e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, 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(5.9995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8207e-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(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8135e-05, 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(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2526e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5869e-05, 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(1.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4772e-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>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7887323943661971, 'agreement': 0.5529411764705883, 'direct_attack': 0.36, 'undercutter_attack': 0.3404255319148936, 'partial': 0.5384615384615384}, 'recall': {'support': 0.8562691131498471, 'agreement': 0.44339622641509435, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.2962962962962963, 'partial': 0.2916666666666667}, 'f1': {'support': 0.8211143695014662, 'agreement': 0.49214659685863876, 'direct_attack': 0.4044943820224719, 'undercutter_attack': 0.3168316831683168, 'partial': 0.3783783783783784}, 'support': {'support': 327, 'agreement': 106, 'direct_attack': 39, 'undercutter_attack': 54, 'partial': 24}, 'micro_avg': {'precision': 0.6690909090909091, 'recall': 0.6690909090909091, 'f1': 0.6690909090909091, 'support': None}, 'macro_avg': {'precision': 0.5161121282426435, 'recall': 0.4698333528132732, 'f1': 0.48259308198585443, 'support': None}, 'weighted_avg': {'precision': 0.6579514787456545, 'recall': 0.6690909090909091, 'f1': 0.6593403835728033, '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
				 {'precision': {'support': 0.0, 'agreement': 0.10294117647058823, 'direct_attack': 0.06310679611650485, 'undercutter_attack': 0.11637931034482758, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 0.1875, 'direct_attack': 0.2708333333333333, 'undercutter_attack': 0.4153846153846154, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.1329113924050633, 'direct_attack': 0.10236220472440943, 'undercutter_attack': 0.18181818181818182, 'partial': 0.0}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.09501557632398754, 'recall': 0.09501557632398754, 'f1': 0.09501557632398754, 'support': None}, 'macro_avg': {'precision': 0.05648545658638413, 'recall': 0.17474358974358975, 'f1': 0.08341835578953091, 'support': None}, 'weighted_avg': {'precision': 0.03445980241543911, 'recall': 0.09501557632398754, 'f1': 0.04924866603476723, 'support': None}}
Loss: tensor(2.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1196, 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
				 {'precision': {'support': 0.8538812785388128, 'agreement': 0.918918918918919, 'direct_attack': 0.34782608695652173, 'undercutter_attack': 0.47580645161290325, 'partial': 0.45}, 'recall': {'support': 0.9765013054830287, 'agreement': 0.30357142857142855, 'direct_attack': 0.16666666666666666, 'undercutter_attack': 0.9076923076923077, 'partial': 0.2647058823529412}, 'f1': {'support': 0.9110840438489647, 'agreement': 0.4563758389261745, 'direct_attack': 0.22535211267605634, 'undercutter_attack': 0.6243386243386243, 'partial': 0.33333333333333337}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7538940809968847, 'recall': 0.7538940809968847, 'f1': 0.7538940809968847, 'support': None}, 'macro_avg': {'precision': 0.6092865472054314, 'recall': 0.5238275181532746, 'f1': 0.5100967906246306, 'support': None}, 'weighted_avg': {'precision': 0.7677235516013021, 'recall': 0.7538940809968847, 'f1': 0.7208590780026162, 'support': None}}
Loss: tensor(1.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8756, 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(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6373, 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(2.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, 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(2.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033, 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(5.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9727, 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(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0907, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4463, 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(1.9224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7405, 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(0.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3966, 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(1.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5897, 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(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0616, 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
				 {'precision': {'support': 0.8298850574712644, 'agreement': 0.7049180327868853, 'direct_attack': 0.26, 'undercutter_attack': 0.7142857142857143, 'partial': 0.2647058823529412}, 'recall': {'support': 0.9425587467362925, 'agreement': 0.38392857142857145, 'direct_attack': 0.2708333333333333, 'undercutter_attack': 0.3076923076923077, 'partial': 0.5294117647058824}, 'f1': {'support': 0.882640586797066, 'agreement': 0.4971098265895954, 'direct_attack': 0.2653061224489796, 'undercutter_attack': 0.4301075268817205, 'partial': 0.35294117647058826}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7087227414330218, 'recall': 0.7087227414330218, 'f1': 0.7087227414330217, 'support': None}, 'macro_avg': {'precision': 0.554758937379361, 'recall': 0.48688494477927746, 'f1': 0.48562104783758997, 'support': None}, 'weighted_avg': {'precision': 0.7238401372464126, 'recall': 0.7087227414330218, 'f1': 0.6953572094177162, 'support': None}}
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5514, 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(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0037, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5030, 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(0.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, 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(0.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0299, 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(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, 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(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8420, 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(1.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, 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(0.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, 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(1.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2115, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, 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(0.5413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, 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(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, 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(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4055, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, 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(4.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, 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(1.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.2537, 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(0.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, 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(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, 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(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2849, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, 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(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, 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(1.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6138, 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(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.3682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, 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(4.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, 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(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9224, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.6306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, 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(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7700, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, 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(0.0402, 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(0.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, 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(7.2945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4668, 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(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2976, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7813, 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(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1789, 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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, 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(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0227, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(1.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3253, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091, 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(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9624, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, 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(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5401, 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(0.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, 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(2.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, 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(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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
				 {'precision': {'support': 0.891358024691358, 'agreement': 0.6818181818181818, 'direct_attack': 0.3090909090909091, 'undercutter_attack': 0.5512820512820513, 'partial': 0.23684210526315788}, 'recall': {'support': 0.9425587467362925, 'agreement': 0.4017857142857143, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.6615384615384615, 'partial': 0.2647058823529412}, 'f1': {'support': 0.9162436548223352, 'agreement': 0.5056179775280899, 'direct_attack': 0.33009708737864074, 'undercutter_attack': 0.6013986013986014, 'partial': 0.25}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7398753894080997, 'recall': 0.7398753894080997, 'f1': 0.7398753894080997, 'support': None}, 'macro_avg': {'precision': 0.5340782544291316, 'recall': 0.5249510943160153, 'f1': 0.5206714642255335, 'support': None}, 'weighted_avg': {'precision': 0.7421745924751881, 'recall': 0.7398753894080997, 'f1': 0.7336232127183556, 'support': None}}
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4798, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, 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(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1860, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2603, 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(0.0084, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, 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(2.8021, 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(0.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0020, 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(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, 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(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, 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(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, 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(4.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9374, 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(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, 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(0.0022, 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(0.0907, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, 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(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, 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(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, 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(0.2663, 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(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, 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(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5123, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, 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(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, 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(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, 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(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, 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(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, 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(2.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, 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(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, 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(1.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.9822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, 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(0.0554, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1366, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, 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(0.0023, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2682, 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(0.0010, 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(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1574, 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(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0561, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, 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(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, 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(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0020, 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(0.0444, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, 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(0.0004, 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(0.0014, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, 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(0.0003, 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(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, 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(0.0043, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, 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(0.0010, 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(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, 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(0.0239, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, 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(1.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7392, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, 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(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1141, 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(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, 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(2.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, 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(0.0029, 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(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, 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(0.0035, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3276, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, 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(1.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5777, 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(0.0038, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, 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(0.0400, 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(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8714, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, 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(1.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5633, 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(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4702, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6089, 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(0.1150, 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(0.0030, 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(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, 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(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, 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(0.1995, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5380, 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(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2878, 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(0.0019, 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(0.2806, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, 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(4.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, 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(0.0125, 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(0.0674, 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(0.0932, 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(0.0458, 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(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, 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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5608, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, 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(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, 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(0.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5449, 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(0.0121, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3118, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.1026, 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(0.1949, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, 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(1.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, 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(0.0016, 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(0.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, 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(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6516, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(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(0.0043, 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(2.8243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3043, 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(0.0020, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, 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(0.0045, 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(0.6196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, 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
				 {'precision': {'support': 0.8351648351648352, 'agreement': 0.7380952380952381, 'direct_attack': 0.06666666666666667, 'undercutter_attack': 0.49137931034482757, 'partial': 0.5714285714285714}, 'recall': {'support': 0.9921671018276762, 'agreement': 0.2767857142857143, 'direct_attack': 0.020833333333333332, 'undercutter_attack': 0.8769230769230769, 'partial': 0.23529411764705882}, 'f1': {'support': 0.9069212410501194, 'agreement': 0.40259740259740256, 'direct_attack': 0.031746031746031744, 'undercutter_attack': 0.6298342541436464, 'partial': 0.3333333333333333}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7429906542056075, 'recall': 0.7429906542056075, 'f1': 0.7429906542056075, 'support': None}, 'macro_avg': {'precision': 0.5405469243400278, 'recall': 0.4804006688033719, 'f1': 0.4608864525741067, 'support': None}, 'weighted_avg': {'precision': 0.7119984815199124, 'recall': 0.7429906542056075, 'f1': 0.6950749435974839, 'support': None}}
Loss: tensor(0.0073, 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(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, 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(0.0035, 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(0.1832, 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(0.0021, 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(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, 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(0.0009, 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(0.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0306, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, 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(0.0084, 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(0.0047, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0003, 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(0.0105, 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(0.0186, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0009, 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(0.0002, 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(0.0002, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, 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(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9770e-05, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(1.4306, 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(0.0079, 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(0.0095, 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(0.0007, 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(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2338, 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(0.0043, 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(0.0101, 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(0.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1606, 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(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, 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(0.0015, 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(1.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, 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(0.0036, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, 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(0.0094, 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(0.0111, 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(0.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8519, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, 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(0.0207, 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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(1.3772, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0004, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, 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(0.0218, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, 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(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, 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(0.0529, 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(0.0039, 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(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0013, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, 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(0.0003, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3426, device='cuda:0', grad_fn=<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(0.0838, 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(0.0004, 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(0.0041, 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(0.0040, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1398, 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(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, 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(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9223, 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(0.0094, 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(0.0098, 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(0.0039, 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(0.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, 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(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, 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(0.0245, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, 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(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0651, 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(0.0210, 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(0.0027, 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(0.0944, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4169, 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(0.1128, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, 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(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5149, 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(0.0070, 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(0.0005, 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(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, 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(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, 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(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, 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(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6371, device='cuda:0', grad_fn=<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(0.0001, 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(0.0150, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0021, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, 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(0.0248, 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(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, 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(0.0033, 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(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4640, 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(0.0446, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, 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(0.0003, 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(0.0010, 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(0.0329, 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(0.0216, 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(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4232, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0005, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, 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(0.0117, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, 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(0.0006, 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(0.0817, 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(0.0005, 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(0.0024, 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(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, 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(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, 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(0.0107, 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(0.0205, 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(0.0935, 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(0.0025, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0002, 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(0.0008, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0019, 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(0.1644, 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(0.0059, 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(0.0003, 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(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, 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(0.0004, 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(1.3602, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, 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(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, 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(0.0003, 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(0.6808, 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(1.2719, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, 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(0.0398, 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(0.0066, 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(0.0188, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, 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(0.0020, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, 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(0.0596, 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(2.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, 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(0.7167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, 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(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, 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(0.3907, device='cuda:0', grad_fn=<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(0.0059, 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(0.0368, 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(0.0109, 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(0.0113, 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(0.0029, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0103, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, 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(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3485, 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(0.0043, 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(2.8187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, 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(0.0004, 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(0.0056, 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(0.0013, 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(0.0004, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, 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(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8929, 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(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, 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(0.0197, 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(0.0024, 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(0.0355, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, 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(0.0117, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8618, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(8.3109e-05, 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(0.0056, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.0004, 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(1.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, 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(0.0001, 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(0.0185, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, 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(0.0065, 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(0.5687, 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(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8398, 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(0.5578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, 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(0.1427, 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(0.0096, 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(0.0093, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, 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(0.0014, 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(1.1503, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4107, 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(0.0005, 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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, 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(0.7210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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
				 {'precision': {'support': 0.8532110091743119, 'agreement': 0.6808510638297872, 'direct_attack': 0.34375, 'undercutter_attack': 0.5333333333333333, 'partial': 0.25}, 'recall': {'support': 0.9712793733681462, 'agreement': 0.2857142857142857, 'direct_attack': 0.22916666666666666, 'undercutter_attack': 0.6153846153846154, 'partial': 0.38235294117647056}, 'f1': {'support': 0.9084249084249083, 'agreement': 0.4025157232704402, 'direct_attack': 0.27499999999999997, 'undercutter_attack': 0.5714285714285715, 'partial': 0.3023255813953488}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7289719626168224, 'recall': 0.7289719626168224, 'f1': 0.7289719626168223, 'support': None}, 'macro_avg': {'precision': 0.5322290812674865, 'recall': 0.49677957646203685, 'f1': 0.4919389569038538, 'support': None}, 'weighted_avg': {'precision': 0.7207193182700379, 'recall': 0.7289719626168224, 'f1': 0.7065894514693585, 'support': None}}
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(0.0024, 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(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0006, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, 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(0.0008, 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(0.0098, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, 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(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0012, 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(0.0116, 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(0.0006, 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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, 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(0.1323, 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(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2429, 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(0.0042, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0045, 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(0.0022, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(0.0003, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, 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(0.0129, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0062, 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(0.0007, 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(0.0004, 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(0.0035, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0001, 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(0.0024, 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(0.0023, 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(0.0156, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0088, 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(0.0174, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, 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(0.0003, 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(0.0002, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4247, device='cuda:0', grad_fn=<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(0.0001, 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(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<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(0.0673, 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(0.0008, 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(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, 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(0.0413, 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(0.0096, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, 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(0.0163, 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(0.0101, 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(0.0564, 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(0.0025, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6640e-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(0.0004, 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(0.0067, 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(9.6036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(9.1828e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0178, 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(0.0110, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0082, 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(0.0212, 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(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0001, 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(0.0049, 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(0.0016, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0232e-05, 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(0.0018, 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(0.0065, 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(0.0169, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, 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(0.0429, 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(0.0007, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, 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(0.0038, 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(0.1706, 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(0.0139, 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(0.0001, 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(0.0071, 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(0.0014, 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(0.0001, 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(0.0065, 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(0.0093, 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(0.0101, 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(0.0558, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4359e-05, 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(0.0042, 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(0.0086, 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(0.0004, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0080, 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(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0003, 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(0.0070, 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(0.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, 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(0.0093, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9474e-05, 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(0.0628, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0003, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0563, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, 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(0.0002, 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(0.0507, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0001, 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(0.8452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.1532, 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(0.0018, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0046, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, 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(0.0309, 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(0.0018, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0008, 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(0.0026, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, 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(0.0327, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0136, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0046, 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(0.0004, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.0054, 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(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(8.6561e-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(0.0032, 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(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0004, 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(0.2172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0009, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0040, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(7.6086e-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(0.0053, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0014, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7520, 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(0.0128, 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(0.0334, 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(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7126e-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(0.0038, 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(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, 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(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-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(5.4106e-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(5.9192e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6853e-05, 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(0.0003, 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(0.0074, 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(0.1486, 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(0.0041, 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(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5068e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, 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(0.0022, 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(0.0033, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0018, 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(0.0063, device='cuda:0', grad_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(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-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(0.0567, 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(0.0027, 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(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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8264e-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(0.0037, 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(0.0007, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6642e-05, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(6.1039e-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(0.0020, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, 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(8.2352e-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(7.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4985e-05, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8494, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0011, 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(0.0003, 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(0.0058, 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(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0018, 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(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(0.0055, 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(6.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-05, 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(0.0002, 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(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8478, 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(0.0096, 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(8.5501e-05, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, 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(0.0260, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9155e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0057, 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(0.2346, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0114, 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(0.0044, 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(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3913e-05, 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(0.0032, 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(0.0343, 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(0.0303, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, 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(0.0605, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1837e-05, 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(0.1522, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1494, 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(0.0187, 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(0.0028, 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(0.0008, 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(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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, 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(0.0057, 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(0.0017, 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(0.0084, 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(0.0003, 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(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(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1712, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, 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(0.0018, 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(9.0101e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, 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(0.0107, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0052, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, 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(0.0032, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, 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(0.0076, 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(0.0356, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_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(0.0005, 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(0.0003, 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(0.1346, 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(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, 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(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0047, 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(0.0047, 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(0.0130, 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(0.0005, 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(0.0009, 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(0.0034, 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(0.0050, 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(0.0004, 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(0.0097, 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(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, 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(0.0042, 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(0.0012, 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(0.0241, 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(0.0007, 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(0.0021, 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(0.0006, 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(0.0025, 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(0.0018, 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(0.0007, 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(0.0050, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, 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(0.0056, 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(0.0083, 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(0.0003, 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(0.0126, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(0.6218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(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(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4429e-05, 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(0.0547, 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(5.0560, 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(0.0418, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0025, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, 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(0.0182, 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(0.0002, 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(0.0002, 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(0.4784, 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(6.8063e-05, 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(0.0064, 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(5.1805e-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(0.0050, device='cuda:0', grad_fn=<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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8961e-05, 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(0.0019, 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(0.0026, 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(8.3593e-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(0.0066, 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
				 {'precision': {'support': 0.8407079646017699, 'agreement': 0.6607142857142857, 'direct_attack': 0.32075471698113206, 'undercutter_attack': 0.576271186440678, 'partial': 0.2727272727272727}, 'recall': {'support': 0.9921671018276762, 'agreement': 0.33035714285714285, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5230769230769231, 'partial': 0.17647058823529413}, 'f1': {'support': 0.9101796407185629, 'agreement': 0.44047619047619047, 'direct_attack': 0.33663366336633666, 'undercutter_attack': 0.5483870967741935, 'partial': 0.21428571428571427}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7383177570093458, 'recall': 0.7383177570093458, 'f1': 0.7383177570093457, 'support': None}, 'macro_avg': {'precision': 0.5342350852930278, 'recall': 0.4752476845327406, 'f1': 0.4899924611241996, 'support': None}, 'weighted_avg': {'precision': 0.7135790206369836, 'recall': 0.7383177570093458, 'f1': 0.7118713818476073, 'support': None}}
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, 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(6.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, 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(0.0001, 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(0.0151, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9577e-05, 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(0.0010, 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(0.0002, 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(0.0147, 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(0.0005, 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(0.0057, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0008, 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(3.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, 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(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0099, 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(0.0063, 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(0.0024, 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(0.0489, device='cuda:0', grad_fn=<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(0.0073, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, 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(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, 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(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, 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(0.0203, 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(0.0038, 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(0.0004, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1523, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, 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(0.0174, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0014, 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(0.0920, 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(0.0002, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2834, 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(0.0065, device='cuda:0', grad_fn=<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(0.0109, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1956, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, 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(0.0020, 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(0.0783, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9697e-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(0.0075, 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(0.0090, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, 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(0.0016, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-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(0.0013, 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(0.0033, 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(0.0055, 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(0.0007, 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(0.0002, 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(0.0019, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0553, 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(0.0009, 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(5.2350e-05, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0005, 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(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7447e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, 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(0.0308, 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(0.0073, 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(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0004, 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(0.0012, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, 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(0.0048, 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(0.0002, 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(0.0001, 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(0.0002, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8951e-05, 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(0.0190, 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(0.0001, 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(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-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(0.0197, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(3.9301e-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(0.0022, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8456e-05, 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(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0002, 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(0.0016, 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(5.8042e-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(0.0001, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4761e-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(0.0121, 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(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(0.0026, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2292e-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(0.0009, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0033, 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(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8304e-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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1917e-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(0.0010, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0309, device='cuda:0', grad_fn=<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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-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(0.0003, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-05, 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(0.0040, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0022, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-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(0.0332, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7600, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-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(0.0002, 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(0.0023, 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(0.0039, 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(0.0031, 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(0.0029, 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(0.0025, 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(8.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, 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(8.1323e-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(0.0054, 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(0.0039, 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(0.0007, 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(0.0001, 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(0.0010, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, 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(0.0018, 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(0.0006, 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(9.1433e-05, 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(0.0005, 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(5.8314e-05, 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(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2724e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3884e-05, device='cuda:0', grad_fn=<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(0.0031, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, 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(0.0070, 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(0.0029, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7650e-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(0.0359, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, 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(0.3437, 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(3.3821e-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(0.0005, 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(0.0009, 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(2.9037e-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(0.0037, 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(0.0038, 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(3.4245e-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(7.3512e-05, 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(0.0013, 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(0.0059, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4963e-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(7.9204e-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(0.0005, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, 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(0.0087, 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(9.4944e-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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-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(0.0139, 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(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, 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(0.0002, 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(0.0003, 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(0.0023, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7254e-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(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-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(0.0009, 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(0.0022, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, 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(0.0022, 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(0.0022, 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(0.0005, 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(0.3369, 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(3.1550e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, 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(0.0020, 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(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(0.0042, 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(0.0013, 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(0.0010, 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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2489e-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(0.0017, 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(0.0059, 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(0.0001, 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(0.0111, 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(0.0025, 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(0.5057, 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(0.0019, 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(0.0137, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0062, 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(0.0091, 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(0.0009, 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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-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(0.0396, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0393e-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(0.0212, 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(0.0186, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, 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(0.0521, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0032, 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(5.4318e-05, device='cuda:0', grad_fn=<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(0.0039, 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(5.0170e-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(6.6125e-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(0.0027, 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(0.0008, 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(5.2743e-05, 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(0.0028, 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(4.2752e-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(0.0023, 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(0.0010, 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(0.0020, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, 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(0.5231, 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(0.0013, 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(0.0033, 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(0.0002, 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(0.0002, 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(0.0024, 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(0.0004, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, 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(0.0026, 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(0.0009, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2849, 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(0.0016, 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(0.0007, 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(0.0069, 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(0.0022, 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(0.0017, 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(0.0464, 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(0.0015, 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(0.0024, 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(0.0018, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, 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(0.0029, 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(0.0044, 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(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1332e-05, 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(0.0017, 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(0.0036, 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(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7336e-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(0.0007, 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(0.0021, 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(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, 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(0.0002, 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(3.7575e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2400, device='cuda:0', grad_fn=<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(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0011, 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(0.0001, 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(5.0412e-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(0.0011, 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(0.0019, 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(0.0034, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, 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(0.0006, 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(0.0089, 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(5.1835e-05, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0210, 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(4.2177e-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(0.0053, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(0.0190, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, 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(0.0035, 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(0.0070, device='cuda:0', grad_fn=<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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-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(0.0023, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8305e-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(0.0030, 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(0.0007, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5464, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0060, 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(0.0001, 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(0.0033, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8787e-05, 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(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9677e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-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(5.8617e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(6.2765e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4793, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, 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(9.4795e-05, 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(0.0332, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5037e-05, 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(8.1656e-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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, 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(0.0286, 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
				 {'precision': {'support': 0.8403547671840355, 'agreement': 0.7777777777777778, 'direct_attack': 0.3076923076923077, 'undercutter_attack': 0.5526315789473685, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.25, 'direct_attack': 0.25, 'undercutter_attack': 0.6461538461538462, 'partial': 0.29411764705882354}, 'f1': {'support': 0.9088729016786571, 'agreement': 0.3783783783783784, 'direct_attack': 0.27586206896551724, 'undercutter_attack': 0.5957446808510639, 'partial': 0.27027027027027023}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7336448598130841, 'recall': 0.7336448598130841, 'f1': 0.7336448598130842, 'support': None}, 'macro_avg': {'precision': 0.5456912863202978, 'recall': 0.48596552579658087, 'f1': 0.4858256600287773, 'support': None}, 'weighted_avg': {'precision': 0.7292169319990754, 'recall': 0.7336448598130841, 'f1': 0.703474567719871, 'support': None}}
Loss: tensor(0.0041, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3908, 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(0.0021, 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(0.0016, 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(0.0015, 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(7.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, 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(0.0015, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, 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(8.6408e-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(7.6661e-05, 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(0.0001, 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(6.5065e-05, 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>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2097e-05, 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(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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5287e-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(0.0004, 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(0.0005, 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(8.2655e-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(9.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-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(0.0032, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2744e-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(0.0011, 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(0.0031, 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(0.0056, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6307e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2473, 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(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0068, 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(0.0007, 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(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8163e-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(8.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8315e-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(0.0095, 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(0.0093, 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(0.0058, 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(6.7094e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-05, 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(0.0021, 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(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0006, 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(0.0058, 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(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9649e-05, 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(0.0010, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1535e-05, 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(0.0078, 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(0.0001, 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(0.0095, 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(0.0002, 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(0.0026, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-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(0.0003, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-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(0.0071, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3109e-05, 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(0.0004, 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(0.0050, 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(0.0006, 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(5.4984e-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(0.0030, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, 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(3.9543e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0051, 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(0.2129, 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(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2532e-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(0.0002, 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(0.0051, 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(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-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(0.0124, 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(0.0016, 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(0.0079, 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(0.0019, 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(8.3592e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0061e-05, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0878e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0626e-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(0.0027, 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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3857e-05, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(1.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5874e-05, 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(0.0022, 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(0.0009, 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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2855e-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(0.0007, 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(0.0001, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8002, 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(0.0017, 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(0.0007, 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(3.7393e-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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, 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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, 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(0.0064, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, 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(0.0029, 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(0.0154, 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(0.0028, 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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-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(0.0001, 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(3.4699e-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(0.0004, 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(0.0040, 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(0.0002, 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(4.0330e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.0450, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9851e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, 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(9.6760e-05, 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(0.0003, 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(0.0001, 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(0.0018, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1049e-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(0.0001, 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(8.8679e-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(8.1505e-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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-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(0.0001, 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(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-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(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4864e-05, 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(4.8202e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8588e-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(0.0008, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8317e-05, 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(0.0008, 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(0.0006, 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(4.3691e-05, 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(0.0010, 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(4.4326e-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(8.0535e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-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(0.0023, 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(0.0046, 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(0.0015, 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(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0066, 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(7.5087e-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(0.0008, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, 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(7.9537e-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(0.0008, 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(0.0005, 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(0.0155, 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(4.6022e-05, 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(0.0009, 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(0.0023, 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(4.5931e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0002, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3491e-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(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-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(0.0105, 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(0.0047, 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(3.1853e-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(6.6426e-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(8.1081e-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(7.1212e-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(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5186e-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(0.0012, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5187e-05, 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(6.2007e-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(0.0064, 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(0.0022, 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(0.0021, 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(0.0005, 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(7.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5881e-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(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6287e-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(5.8011e-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(4.5568e-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(0.0003, 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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(7.9537e-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(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0001, 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(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0012, 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(0.0149, 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(0.0014, 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(0.0011, 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(0.0009, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-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(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, 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(0.0073, 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(0.0143, 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(9.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8485e-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(0.0013, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9362e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-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(0.0010, 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(0.0005, 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(4.6294e-05, 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(8.1290e-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(0.0009, 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(0.0041, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-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(0.0045, 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(0.0100, 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(0.0045, 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(6.9486e-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(5.7042e-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(0.0046, 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(5.4287e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-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(0.0028, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, 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(7.4268e-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(0.0002, 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(0.0005, 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(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1101e-05, 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(0.0298, 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(0.0020, 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(0.0003, 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(0.0005, 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(4.6446e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6003e-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(9.6457e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0001, 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(0.0014, 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(6.2037e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(6.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0799e-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(0.0021, 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(3.4850e-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(5.1018e-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(0.0015, 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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4610e-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(0.0025, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5054e-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(6.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, 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(0.0043, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, 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(0.0007, 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(0.0052, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6702e-05, 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(0.0018, 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(0.0003, 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(8.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<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(0.0049, 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(0.0013, 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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0352e-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(0.0024, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2219e-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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6426e-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(0.0003, 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(0.0026, 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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4874e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, 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(0.0007, 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(9.6848e-05, 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(0.0009, 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(0.0034, 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(0.0004, 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(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(3.4668e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0023, 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(8.2442e-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(3.4578e-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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7660e-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(0.0005, 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(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(0.0010, 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(0.0060, 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(0.0001, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<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(2.5070e-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(0.2844, 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(0.6355, 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(5.0836e-05, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, 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(4.9927e-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(0.0003, 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(0.0013, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-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(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, 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(0.0004, 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(2.4859e-05, 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(0.0005, 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
				 {'precision': {'support': 0.8627002288329519, 'agreement': 0.6470588235294118, 'direct_attack': 0.3584905660377358, 'undercutter_attack': 0.5757575757575758, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9843342036553525, 'agreement': 0.39285714285714285, 'direct_attack': 0.3958333333333333, 'undercutter_attack': 0.5846153846153846, 'partial': 0.14705882352941177}, 'f1': {'support': 0.9195121951219511, 'agreement': 0.4888888888888888, 'direct_attack': 0.37623762376237624, 'undercutter_attack': 0.5801526717557253, 'partial': 0.1923076923076923}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7523364485981309, 'recall': 0.7523364485981309, 'f1': 0.7523364485981309, 'support': None}, 'macro_avg': {'precision': 0.5443569943870906, 'recall': 0.500939777598125, 'f1': 0.5114198143673268, 'support': None}, 'weighted_avg': {'precision': 0.7273535980012663, 'recall': 0.7523364485981309, 'f1': 0.7308980022281005, 'support': None}}
Loss: tensor(0.0003, 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(0.0042, 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(0.0095, 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(0.0003, 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(0.0023, device='cuda:0', grad_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(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-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(5.9252e-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(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0024, 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(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0071, 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(0.0022, 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(0.0005, 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(6.9395e-05, 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(0.0017, 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(0.0001, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3057e-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(0.0004, 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(0.0004, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-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(2.9552e-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(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(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1321e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-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(7.0848e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8425e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<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(0.0008, 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(5.1987e-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(9.3672e-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(0.0018, 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(0.0001, 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(0.0023, 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(0.0003, 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(0.0001, 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(6.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5237e-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(0.0021, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-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(7.1633e-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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6498e-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(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8900e-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(0.0011, 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(4.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2483e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2292e-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(6.6489e-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(0.0013, 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(8.0626e-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(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(0.0017, 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(0.0004, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-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(0.0021, 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(0.0002, 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(0.0002, 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(6.1826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9739e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(4.5780e-05, 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(6.9637e-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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-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(0.0019, 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(9.0676e-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(0.0015, 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(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1704e-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(4.7112e-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(0.0004, 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(9.6884e-05, 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(9.0159e-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(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(6.9031e-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(0.0002, 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(5.9101e-05, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(7.3996e-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(3.7424e-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(0.0003, 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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-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(0.0141, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2007e-05, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8516e-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(0.0004, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, 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(0.0009, 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(0.0003, 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(0.0031, 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(0.0064, 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(0.0010, 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(0.0008, 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(8.0172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-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(4.1662e-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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2270e-05, 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(0.0004, 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(0.0017, 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(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-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(0.0006, 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(0.0016, 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(6.6669e-05, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1483e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9010e-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(0.0031, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2753e-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(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8063e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9727e-05, 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(0.0029, 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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-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(0.0017, 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(0.0001, 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(9.1702e-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(0.0188, 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(0.0009, 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(0.0033, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-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(4.9171e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-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(0.0008, 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(0.0004, 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(0.0046, 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(0.0048, 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(0.0006, 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(0.0072, 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(2.6342e-05, 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(9.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, 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(0.0001, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-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(0.0011, 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(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0040e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1999, 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(0.0008, 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(0.0021, 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(3.1429e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5367e-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(8.4710e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, 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(0.0018, 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(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-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(0.0016, 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(5.6104e-05, 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(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(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-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(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1733e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0012, 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(4.6264e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(3.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-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(8.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0120e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0038, 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(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-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(7.9384e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9273e-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(0.0027, 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(0.0007, 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(3.3851e-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(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<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(7.9719e-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(8.3894e-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(0.0008, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-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(0.0018, 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(0.0016, 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(0.0020, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-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(0.0011, 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(2.4435e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-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(0.0021, 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(8.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6516e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-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(3.3367e-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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0010, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3852e-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(0.0043, 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(0.0029, 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(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-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(0.0014, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, 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(3.8332e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-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(7.8900e-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(3.2549e-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(0.0032, 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(2.6251e-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(0.0040, 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(0.0006, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0001, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(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(0.0024, 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(0.0012, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-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(0.0001, 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(0.0011, 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(0.0035, 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(0.0047, 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(0.0004, 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(0.0001, 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(0.0005, 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(0.0041, 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(0.0026, 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(0.0018, 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(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(6.4034e-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(2.4586e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-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(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, 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(7.0151e-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(0.0015, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-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(0.0025, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(7.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2977e-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(8.2807e-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(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4734e-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(0.0010, 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(0.0005, 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(0.0006, 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(0.0006, 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(0.0019, 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(0.0019, 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(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(0.0008, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, 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(0.0006, 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(5.7708e-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(0.0003, 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(0.0008, 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(0.0029, 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(9.6400e-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(6.5247e-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(0.0013, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-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(5.7949e-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(3.8514e-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(7.2026e-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(0.0004, 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(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-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(0.0007, 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(3.9875e-05, 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(0.0001, 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(0.0004, 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(0.0007, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7034e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-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(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(6.9696e-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(0.0005, 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(0.0017, 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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2863e-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(3.8150e-05, device='cuda:0', grad_fn=<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(6.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, 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(0.0003, 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(7.7266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-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(0.0006, 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(0.0008, 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(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, 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(3.1731e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, 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(7.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4356e-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(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2633e-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(0.0026, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6979e-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(0.0006, 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(0.0007, 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(0.0001, 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(0.0003, 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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.68, 'direct_attack': 0.3508771929824561, 'undercutter_attack': 0.5714285714285714, 'partial': 0.2608695652173913}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.30357142857142855, 'direct_attack': 0.4166666666666667, 'undercutter_attack': 0.5538461538461539, 'partial': 0.17647058823529413}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4197530864197531, 'direct_attack': 0.38095238095238093, 'undercutter_attack': 0.5625, 'partial': 0.2105263157894737}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7398753894080997, 'recall': 0.7398753894080997, 'f1': 0.7398753894080997, 'support': None}, 'macro_avg': {'precision': 0.5414546650348151, 'recall': 0.4880221946179556, 'f1': 0.49695789509386, 'support': None}, 'weighted_avg': {'precision': 0.7200997818027928, 'recall': 0.7398753894080997, 'f1': 0.7133232879367832, 'support': None}}
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(9.9574e-05, device='cuda:0', grad_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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0081e-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(2.3193e-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(0.0009, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-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(9.2706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-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(0.0017, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4206e-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(0.0004, 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(0.0019, 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(0.0013, 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(2.4949e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, 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(0.0002, 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(0.0008, 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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5024e-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(0.0016, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-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(0.0009, 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(0.0003, 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(4.4326e-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(0.0022, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9152e-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(0.0004, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-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(0.0009, 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(6.5187e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8778e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3734e-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(4.2964e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6022e-05, device='cuda:0', grad_fn=<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(6.0736e-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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-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(3.1852e-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(0.0010, 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(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5820e-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(6.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7163e-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(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(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0007, 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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-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(0.0001, 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(0.0007, 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(6.5730e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-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(9.3572e-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(0.0004, 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(7.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5086e-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(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-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(4.2026e-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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-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(0.0021, 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(0.0002, 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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-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(0.0092, device='cuda:0', grad_fn=<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(0.0011, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-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(0.0028, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4359e-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(0.0031, 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(4.7081e-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(0.0023, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, 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(5.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3975e-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(0.0002, 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(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0283e-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(0.0005, 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(9.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, 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(0.0004, 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(8.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-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(0.0044, 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(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-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(0.6892, 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(6.1916e-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(0.0007, 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(9.1007e-05, 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(7.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-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(0.0010, 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(9.5489e-05, device='cuda:0', grad_fn=<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(4.6597e-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(4.3963e-05, device='cuda:0', grad_fn=<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(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-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(0.0005, 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(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-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(6.2794e-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(0.0007, 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(0.0006, 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(6.9728e-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(2.3314e-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(9.9968e-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(4.3085e-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(0.0015, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-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(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(3.2731e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1119e-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(0.0004, 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(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4478e-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(0.0010, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4184e-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(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-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(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8334e-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(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-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(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(0.0004, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, 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(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(0.0025, 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(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6527e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7297e-05, 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(0.0009, 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(4.5629e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-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(0.0001, 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(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3714, 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(8.6015e-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(4.5023e-05, device='cuda:0', grad_fn=<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(7.9777e-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(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(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(0.0003, 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(0.0012, 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(0.0005, 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(0.0011, 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(0.0005, 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(4.2843e-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(2.7614e-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(0.0015, 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(1.0912, device='cuda:0', grad_fn=<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(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, 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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-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(0.0004, 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(0.0012, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-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(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(0.0003, 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(2.9945e-05, 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(0.0008, 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(0.0023, 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(0.0026, 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(0.0004, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0021, 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(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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-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(0.0003, 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(6.1947e-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(0.0010, 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(0.0013, 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(0.0003, 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(7.6630e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0003, 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(0.0017, 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(0.0001, 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(9.8152e-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(0.0003, 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(2.9824e-05, 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(5.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7668e-05, 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(0.0008, 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(3.6182e-05, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, 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(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(0.0002, 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(0.0007, 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(2.9733e-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(0.0008, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0978e-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(6.3399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-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(0.0021, 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(8.9037e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<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(2.1558e-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(7.0969e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0032, 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(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(0.0014, 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(0.0004, 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(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-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(0.0003, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-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(0.0028, 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(8.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, 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(5.0291e-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(0.0002, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4278e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-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(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5284e-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(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-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(0.0008, 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(9.3456e-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(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, 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(0.0003, 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(0.0002, 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(6.2977e-05, 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(3.6243e-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(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4508e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0684e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3642e-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(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-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(0.0007, 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(0.0021, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-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(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-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(0.0010, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4277e-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(0.0017, 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(3.8877e-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(0.0007, 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(1.9227e-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(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-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(0.0003, 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(5.7287e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(2.8915e-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(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-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(0.0006, 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(0.0004, 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(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(0.0004, 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(0.0007, 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(2.0135e-05, device='cuda:0', grad_fn=<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(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-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(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(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(3.0278e-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(0.0033, 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(7.3118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-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(0.0008, 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(0.0024, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5894e-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(2.3678e-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(0.0014, 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(0.0001, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-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(0.0089, 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(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4084e-05, 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(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-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(2.6494e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-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(0.0004, 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(2.8219e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6214e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-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(3.4608e-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(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7366e-05, 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(8.6586e-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(3.8816e-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(0.0011, 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(0.0004, 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(0.0001, 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(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(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-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(3.6818e-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(3.8634e-05, 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>)
			------------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
				 {'precision': {'support': 0.8475336322869955, 'agreement': 0.6792452830188679, 'direct_attack': 0.3620689655172414, 'undercutter_attack': 0.5806451612903226, 'partial': 0.2608695652173913}, 'recall': {'support': 0.9869451697127938, 'agreement': 0.32142857142857145, 'direct_attack': 0.4375, 'undercutter_attack': 0.5538461538461539, 'partial': 0.17647058823529413}, 'f1': {'support': 0.9119420989143546, 'agreement': 0.43636363636363634, 'direct_attack': 0.39622641509433965, 'undercutter_attack': 0.5669291338582677, 'partial': 0.2105263157894737}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7429906542056075, 'recall': 0.7429906542056075, 'f1': 0.7429906542056075, 'support': None}, 'macro_avg': {'precision': 0.5460725214661637, 'recall': 0.49523809664456264, 'f1': 0.5043975200040144, 'support': None}, 'weighted_avg': {'precision': 0.7237876384892872, 'recall': 0.7429906542056075, 'f1': 0.7183391082851758, 'support': None}}
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-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(3.8392e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-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(0.0003, 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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9393e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-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(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(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7676e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-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(3.3457e-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(0.0005, 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(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9342e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-05, 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(6.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(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(0.0005, 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(9.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, 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(0.0001, 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(0.0045, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-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(3.6848e-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(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0565e-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(1.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2267e-05, device='cuda:0', grad_fn=<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(2.3981e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-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(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(4.1329e-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(4.1753e-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(0.0008, 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(4.1360e-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(0.0002, 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(0.0035, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-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(7.6901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-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(0.0005, 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(0.0011, 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(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-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(0.0026, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<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(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-05, 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(0.0003, 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(0.0008, 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(0.0014, 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(0.0001, 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(2.5192e-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(0.0004, 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(8.0745e-05, device='cuda:0', grad_fn=<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(0.0025, 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(0.0013, 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(3.5607e-05, device='cuda:0', grad_fn=<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(0.0029, device='cuda:0', grad_fn=<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(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-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(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(0.0009, 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(3.4335e-05, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5860e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2965e-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(3.1974e-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(0.0004, 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(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(6.6458e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-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(0.0002, 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(5.7406e-05, 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(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(0.0006, 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(0.0007, 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(7.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0554e-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(0.0011, 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(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2190e-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(2.3526e-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(0.0011, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, 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(7.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-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(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9575e-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(5.4772e-05, 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(0.0006, 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(0.0015, 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(0.0010, 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(4.7899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-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(0.0002, 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(3.7333e-05, device='cuda:0', grad_fn=<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(9.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4548e-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(1.8349e-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(4.1905e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0003, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(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(9.0094e-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(0.0004, 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(0.0002, 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(0.0012, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5247e-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(7.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<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(3.8695e-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(5.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5679e-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(7.3026e-05, device='cuda:0', grad_fn=<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(0.0029, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.3862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-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(9.4124e-05, device='cuda:0', grad_fn=<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(0.0004, 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(3.8120e-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(6.4398e-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(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1885e-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(0.0009, 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(3.0944e-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(0.0005, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4674, 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(0.0001, 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(3.9301e-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(0.0013, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6647e-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(0.0012, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-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(3.2337e-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(0.0009, 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(0.0022, 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(6.9726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5887e-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(0.0002, 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(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(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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-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(5.8466e-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(1.7562e-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(0.0010, 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(3.1187e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9626e-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(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.4539e-05, device='cuda:0', grad_fn=<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(3.2943e-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(0.0013, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7798e-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(6.9607e-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(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7640e-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(0.0021, 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(0.0022, 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(0.0005, 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(0.0013, 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(4.0542e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<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(0.0005, 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(7.3268e-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(4.0633e-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(5.4225e-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(0.0003, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<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(8.7015e-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(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3126e-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(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-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(0.0005, 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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-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(0.0009, 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(0.0023, 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(0.0004, 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(3.5880e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<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(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-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(7.3842e-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(0.0004, 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(8.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-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(0.0015, 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(5.2047e-05, 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(3.2943e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-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(7.5870e-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(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<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(0.0014, 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(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(0.0008, 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(4.2571e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-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(2.3526e-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(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(5.5195e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7097, device='cuda:0', grad_fn=<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(5.3590e-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(0.0008, 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(2.6978e-05, 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(0.0001, 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(0.0013, 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(0.0001, 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(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3884e-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(0.0004, 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(8.4894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-05, device='cuda:0', grad_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(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6993e-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(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1523e-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(4.6022e-05, device='cuda:0', grad_fn=<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(5.7199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0441e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0013, 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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0623e-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(0.0003, 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(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(8.0774e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0004, 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(0.0002, 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(0.0001, 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(2.0014e-05, device='cuda:0', grad_fn=<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(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(5.1834e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4531e-05, device='cuda:0', grad_fn=<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(3.0187e-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(4.6537e-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(2.7402e-05, 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(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(0.0002, 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(5.6284e-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(0.0004, 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(0.0009, 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(0.0017, 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(0.0028, 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(0.0044, 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(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(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(1.9197e-05, 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(0.0001, 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(2.4192e-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(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-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(0.0005, 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(6.5111e-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(0.0014, 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(0.0025, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, 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(3.7938e-05, 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(0.0004, 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(0.0005, 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(0.0024, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3468e-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(0.0003, 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(1.7774e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-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(9.1373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-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(6.6184e-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(8.6464e-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(4.3479e-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(3.9392e-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(0.0006, 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(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-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(9.3222e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8709e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1007e-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(2.5464e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2170e-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(0.0008, 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(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-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(0.0005, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.84375, 'agreement': 0.6666666666666666, 'direct_attack': 0.33962264150943394, 'undercutter_attack': 0.5873015873015873, 'partial': 0.2916666666666667}, 'recall': {'support': 0.9869451697127938, 'agreement': 0.32142857142857145, 'direct_attack': 0.375, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9097472924187726, 'agreement': 0.43373493975903615, 'direct_attack': 0.3564356435643565, 'undercutter_attack': 0.578125, 'partial': 0.2413793103448276}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7414330218068536, 'recall': 0.7414330218068536, 'f1': 0.7414330218068536, 'support': None}, 'macro_avg': {'precision': 0.5458015124288709, 'recall': 0.4916973726626622, 'f1': 0.5038844372173985, 'support': None}, 'weighted_avg': {'precision': 0.7199627309974912, 'recall': 0.7414330218068536, 'f1': 0.7163636428227651, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(3.0066e-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(8.2744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1826e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-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(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1986e-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(8.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, device='cuda:0', grad_fn=<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(5.1108e-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(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(0.0006, 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(0.0001, 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(0.0003, 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(0.0003, 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(0.0003, 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(0.0018, 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(0.0012, 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(0.0014, 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(4.1692e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6127e-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(7.3754e-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(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9755e-05, device='cuda:0', grad_fn=<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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, 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(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-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(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(0.0004, 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(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(9.9796e-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(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-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(0.0003, 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(3.4456e-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(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, 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(0.0002, 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(5.8714e-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(0.0001, 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(5.9706e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-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(0.0007, 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(1.3020e-05, 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(0.0002, 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(0.0023, 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(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1372e-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(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(9.6815e-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(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-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(0.0001, 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(2.3738e-05, 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(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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-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(7.4752e-05, device='cuda:0', grad_fn=<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(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-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(0.0012, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(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(0.0001, 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(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9321e-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(5.9978e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-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(7.7504e-05, 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(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(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-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(0.0002, 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(2.4162e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-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(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6882e-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(0.0010, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-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(0.0008, 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(5.0351e-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(0.9011, 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(0.0001, 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(3.6152e-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(0.0033, device='cuda:0', grad_fn=<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(0.0017, 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(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9059e-05, 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(9.3278e-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(3.1156e-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(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4458e-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(0.0002, 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(1.5079e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-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(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9639e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-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(0.0006, 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(4.0512e-05, 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(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-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(3.1368e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-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(2.4283e-05, 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(0.0001, 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(0.0002, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-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(0.0003, 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(3.0399e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0011, 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(5.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0014, 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(2.8219e-05, 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(5.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-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(2.1922e-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(2.3647e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-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(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3287e-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(0.0007, 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(8.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9594e-05, device='cuda:0', grad_fn=<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(2.5737e-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(0.0003, 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(7.3873e-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(0.0009, 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(0.0002, 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(0.0007, 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(6.2430e-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(4.4266e-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(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(0.0010, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0009, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-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(3.6545e-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(3.5123e-05, device='cuda:0', grad_fn=<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(0.0002, 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(2.7069e-05, device='cuda:0', grad_fn=<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(0.0006, 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(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-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(6.4792e-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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-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(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(4.2631e-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(0.0002, 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(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, 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(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(0.0003, 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(0.0002, 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(9.2615e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8434e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(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(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(2.1074e-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(0.0007, 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(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-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(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0008, 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(8.3967e-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(2.8552e-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(0.0011, 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(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(3.5153e-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(7.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0014, 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(0.0008, 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(0.0016, 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(0.0006, 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(0.0002, 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(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8222e-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(2.5918e-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(1.8742e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-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(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-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(8.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7123e-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(0.0009, 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(0.0015, 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(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(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9955e-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(5.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-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(0.0007, 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(0.0010, 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(2.2224e-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(0.0001, 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(5.2410e-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(0.0005, 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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7283e-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(4.9565e-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(1.6320e-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(0.0009, 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(7.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4885e-05, 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(0.0005, 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(8.3953e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-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(0.0007, 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(0.0001, 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(5.8647e-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(0.0006, 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(1.5594e-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(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(0.0009, 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(8.5920e-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(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-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(4.3267e-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(0.0006, 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(3.2519e-05, 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(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-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(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-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(0.0004, 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(0.0005, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(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(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-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(0.0012, 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(7.3845e-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(0.0009, 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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6610e-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(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7264e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0003, 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(0.0002, 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(0.0003, 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(0.0004, 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(0.0021, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.7754e-05, device='cuda:0', grad_fn=<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(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7014e-05, 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(0.0005, 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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0023, 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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-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(9.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-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(0.0002, 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(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(0.0001, 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(0.0005, 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(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-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(0.0008, 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(7.8052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, 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(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(7.7478e-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(7.6806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2898e-05, 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(0.0006, 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(8.7196e-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(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(6.3489e-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(1.7259e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6984e-05, device='cuda:0', grad_fn=<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(0.0018, 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(0.0002, device='cuda:0', grad_fn=<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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-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(4.9625e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-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(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0006, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(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(5.5680e-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(4.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-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(6.1626e-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(5.8767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4740e-05, 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(0.0143, device='cuda:0', grad_fn=<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(0.0019, 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(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-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(0.0003, 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(0.0005, 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(5.6165e-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(0.0004, 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(0.0013, 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(8.0081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0986e-05, 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(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-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(3.2398e-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
				 {'precision': {'support': 0.8418708240534521, 'agreement': 0.6730769230769231, 'direct_attack': 0.35185185185185186, 'undercutter_attack': 0.5873015873015873, 'partial': 0.2916666666666667}, 'recall': {'support': 0.9869451697127938, 'agreement': 0.3125, 'direct_attack': 0.3958333333333333, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9086538461538463, 'agreement': 0.426829268292683, 'direct_attack': 0.3725490196078432, 'undercutter_attack': 0.578125, 'partial': 0.2413793103448276}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7414330218068536, 'recall': 0.7414330218068536, 'f1': 0.7414330218068536, 'support': None}, 'macro_avg': {'precision': 0.5491535705900962, 'recall': 0.4940783250436146, 'f1': 0.50550728887984, 'support': None}, 'weighted_avg': {'precision': 0.7208742986405706, 'recall': 0.7414330218068536, 'f1': 0.7157113327392588, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-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(9.1532e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3261e-05, device='cuda:0', grad_fn=<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(1.7138e-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(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3164e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-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(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-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(7.1513e-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(6.3475e-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(1.7804e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-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(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2942e-05, 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(0.0004, 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(8.6050e-05, 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(0.0002, 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(8.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-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(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0899e-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(1.3020e-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(0.0006, 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(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8474e-05, 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(0.0002, device='cuda:0', grad_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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-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(3.2579e-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(1.1415e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<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(8.3298e-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(2.9400e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, 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(0.0002, 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(8.9312e-05, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8838e-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(0.0007, 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(5.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1289e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-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(0.0003, 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(0.0004, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-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(4.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5431e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-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(0.0005, 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(6.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-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(5.8161e-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(0.0007, 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(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<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(4.0300e-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(2.5737e-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(3.8756e-05, 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(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, 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(0.0006, 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(0.0005, 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(0.0004, 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(0.0016, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5338e-05, 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(0.0023, 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(0.0003, 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(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-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(2.2255e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, 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(0.0007, 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(4.7687e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7575e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-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(0.0009, 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(0.0003, 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(0.0011, 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(0.0007, 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(0.0003, 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(0.0002, 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(8.3529e-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(0.0008, 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(0.0002, 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(2.3314e-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(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-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(1.9227e-05, device='cuda:0', grad_fn=<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(1.7229e-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(3.6030e-05, device='cuda:0', grad_fn=<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(0.0003, 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(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1855e-05, device='cuda:0', grad_fn=<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(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-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(0.0004, 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(0.0013, 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(0.0003, 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(0.0012, 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(3.7303e-05, device='cuda:0', grad_fn=<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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-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(2.0559e-05, device='cuda:0', grad_fn=<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(2.8068e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0715e-05, 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(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-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(0.0001, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-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(0.0001, 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(5.6921e-05, 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(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(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(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0573e-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(0.0002, 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(4.7323e-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(7.1804e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-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(0.0004, 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(0.0009, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-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(7.8023e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<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(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-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(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3811e-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(0.0003, 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(0.0003, 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(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4659e-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(0.0003, 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(0.0015, 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(7.5304e-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(0.0002, 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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, 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(3.9876e-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(5.7466e-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(4.3993e-05, device='cuda:0', grad_fn=<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(0.0008, 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(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-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(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7009e-05, device='cuda:0', grad_fn=<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(3.0369e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0753e-05, device='cuda:0', grad_fn=<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(0.0009, 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(7.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-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(0.0006, 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(4.0966e-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(0.0003, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8277e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0429e-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(2.1346e-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(6.1417e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5357e-05, device='cuda:0', grad_fn=<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(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(0.0012, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<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(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4549e-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(0.0009, 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(0.0003, 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(7.4630e-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(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(3.5909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-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(0.0012, 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(0.0002, 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(3.5955e-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(0.0003, 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(2.3799e-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(0.0016, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1915e-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(2.4980e-05, 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(2.8310e-05, device='cuda:0', grad_fn=<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(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, 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(1.8985e-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(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3449e-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(5.1834e-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(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, 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(5.1229e-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(0.0002, 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(5.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-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(0.0010, 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(4.4236e-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(0.0001, 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(0.0004, 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(4.8717e-05, 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(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-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(3.3881e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-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(4.6112e-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(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-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(5.3106e-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(3.5698e-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(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-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(7.5932e-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(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8492e-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(0.0004, 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(2.4768e-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(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(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7922e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-05, 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(8.6053e-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(0.0003, 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(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<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(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-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(3.3790e-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(5.7738e-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(0.0003, 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(4.6263e-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(0.0014, 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(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-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(0.0004, 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(9.5651e-05, 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(8.0443e-05, 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(9.3866e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(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(0.0001, 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(0.0012, 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(0.0007, 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(0.0006, 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(6.6790e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-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(6.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-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(4.3690e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4124e-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(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-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(0.0005, 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(0.0007, 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(1.1678, 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(9.5363e-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(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.2406e-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(1.7440e-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(0.0006, 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(0.0003, 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(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, 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(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4519e-05, 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(3.5970e-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(2.3163e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2604e-05, 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(8.4307e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1291e-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(1.9651e-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(0.0004, 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(4.6385e-05, device='cuda:0', grad_fn=<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(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1798e-05, device='cuda:0', grad_fn=<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(2.2164e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, 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(1.7319e-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(3.9542e-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(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9793e-05, device='cuda:0', grad_fn=<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(8.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6172e-05, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-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(2.0014e-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(0.0005, 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(1.1506e-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(1.9954e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-05, device='cuda:0', grad_fn=<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>)
			------------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
				 {'precision': {'support': 0.8418708240534521, 'agreement': 0.6938775510204082, 'direct_attack': 0.35185185185185186, 'undercutter_attack': 0.59375, 'partial': 0.3076923076923077}, 'recall': {'support': 0.9869451697127938, 'agreement': 0.30357142857142855, 'direct_attack': 0.3958333333333333, 'undercutter_attack': 0.5846153846153846, 'partial': 0.23529411764705882}, 'f1': {'support': 0.9086538461538463, 'agreement': 0.42236024844720493, 'direct_attack': 0.3725490196078432, 'undercutter_attack': 0.5891472868217055, 'partial': 0.26666666666666666}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7429906542056075, 'recall': 0.7429906542056075, 'f1': 0.7429906542056075, 'support': None}, 'macro_avg': {'precision': 0.5578085069236038, 'recall': 0.5012518867759999, 'f1': 0.5118754135394533, 'support': None}, 'weighted_avg': {'precision': 0.7260046552604131, 'recall': 0.7429906542056075, 'f1': 0.7173868600533707, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(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(2.7190e-05, 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(0.0004, 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(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(7.6833e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2754e-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(0.0001, 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(0.0003, 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(0.0010, 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(2.1195e-05, 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(0.0012, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, 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(0.0005, 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(9.2705e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<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(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-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(0.0006, 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(0.0007, 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(5.0835e-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(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(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(6.8183e-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(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8352e-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(4.1904e-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(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0212e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, 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(0.0004, 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(0.0009, 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(2.7432e-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(9.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6811e-05, device='cuda:0', grad_fn=<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(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(0.0003, 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(0.0001, 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(0.0007, 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(2.0317e-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(2.6100e-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(5.0350e-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(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(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, device='cuda:0', grad_fn=<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(7.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4771e-05, 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(3.7726e-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(0.0010, 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(6.7394e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-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(0.0004, 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(0.0011, 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(0.0002, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1522e-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(0.0001, device='cuda:0', grad_fn=<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(0.0018, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5204e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-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(1.9136e-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(7.0410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5662e-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(2.2467e-05, 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(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(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-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(1.3565e-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(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(0.0011, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-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(7.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-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(0.0007, 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(0.0008, 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(3.5001e-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(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-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(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-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(5.6194e-05, device='cuda:0', grad_fn=<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(8.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<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(3.2730e-05, 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(1.7531e-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(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<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(0.0010, 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(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, 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(0.0009, 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(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-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(0.0001, device='cuda:0', grad_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(0.0004, 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(8.9011e-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(2.2951e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8465e-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(9.0608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(6.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, 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(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(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-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(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-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(0.0003, 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(9.3207e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-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(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(7.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7718e-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(0.0005, 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(0.0006, 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(5.4195e-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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<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(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(0.0005, 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(9.4290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2100e-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(2.8159e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7598e-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(9.6177e-05, 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(0.0002, 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(2.0801e-05, device='cuda:0', grad_fn=<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(0.0006, 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(2.2557e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0005, 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(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6280e-05, device='cuda:0', grad_fn=<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(1.4988e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0001, 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(5.7012e-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(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1761e-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(5.2197e-05, 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(5.5708e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4418e-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(5.7618e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-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(1.6683e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7492e-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(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7396e-05, 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(0.0002, 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(0.0001, 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(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7243e-05, device='cuda:0', grad_fn=<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(7.2783e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-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(0.0006, 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(0.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7556e-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(0.0008, 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(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, 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(5.1077e-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(3.6424e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-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(2.9309e-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(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3689e-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(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, 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(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(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-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(0.0006, 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(0.0009, 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(9.9087e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3959e-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(9.8406e-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(8.1664e-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(0.0003, 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(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0005, 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(6.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0008, 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(0.0011, 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(0.0003, 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(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-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(1.1597e-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(1.6048e-05, device='cuda:0', grad_fn=<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(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-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(1.4473e-05, 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(0.0005, 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(2.9400e-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(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(1.4049e-05, device='cuda:0', grad_fn=<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(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-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(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7133e-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(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, 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(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(0.0010, 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(8.2954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5747e-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(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(0.0005, 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(9.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, 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(8.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-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(4.3782e-05, 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(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-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(1.5472e-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(2.3042e-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(4.1934e-05, device='cuda:0', grad_fn=<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(0.0006, 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(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(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, 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(0.0005, 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(1.9136e-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(2.2103e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0371e-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(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4780e-05, 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(0.0003, 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(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-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(0.0003, 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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3508e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7434e-05, 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(2.3466e-05, 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(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(4.8111e-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(0.0005, 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(8.4805e-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(1.3625e-05, 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(4.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9211e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-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(0.4259, device='cuda:0', grad_fn=<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(0.0001, 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(1.8258e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-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(0.0009, 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(1.9075e-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(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-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(0.0006, 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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6563e-05, 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(0.0006, 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(7.0392e-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(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5587e-05, device='cuda:0', grad_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(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(0.0002, 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(5.7648e-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(9.2956e-06, 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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2868e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0720e-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(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(2.0468e-05, 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(6.6370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6690e-05, device='cuda:0', grad_fn=<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(0.0001, 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(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(1.8137e-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(0.0007, 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(2.3920e-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(8.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-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(8.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6911e-05, 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(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(0.0001, 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(0.0013, 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(8.2534e-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(6.5246e-05, device='cuda:0', grad_fn=<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(2.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5929e-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(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4628e-05, 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(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(7.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-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(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1975e-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(1.3807e-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
				 {'precision': {'support': 0.8418708240534521, 'agreement': 0.6938775510204082, 'direct_attack': 0.3508771929824561, 'undercutter_attack': 0.5901639344262295, 'partial': 0.3076923076923077}, 'recall': {'support': 0.9869451697127938, 'agreement': 0.30357142857142855, 'direct_attack': 0.4166666666666667, 'undercutter_attack': 0.5538461538461539, 'partial': 0.23529411764705882}, 'f1': {'support': 0.9086538461538463, 'agreement': 0.42236024844720493, 'direct_attack': 0.38095238095238093, 'undercutter_attack': 0.5714285714285714, 'partial': 0.26666666666666666}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7414330218068536, 'recall': 0.7414330218068536, 'f1': 0.7414330218068536, 'support': None}, 'macro_avg': {'precision': 0.5568963620349707, 'recall': 0.4992647072888204, 'f1': 0.5100123427297341, 'support': None}, 'weighted_avg': {'precision': 0.725568708394329, 'recall': 0.7414330218068536, 'f1': 0.7162211978165858, 'support': None}}
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<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(8.8414e-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(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(0.0005, 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(7.6767e-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(0.0008, 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(0.0002, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3794e-05, device='cuda:0', grad_fn=<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(3.2852e-05, 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(9.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7276e-05, device='cuda:0', grad_fn=<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(1.1718e-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(8.0294e-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(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-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(1.8500e-05, 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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5983e-05, device='cuda:0', grad_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(1.6865e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9976e-05, device='cuda:0', grad_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(2.5494e-05, device='cuda:0', grad_fn=<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(1.3232e-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(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(0.0003, 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(6.8759e-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(5.2425e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-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(1.5018e-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(7.2509e-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(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038e-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(7.2937e-05, 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(2.1740e-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(0.0003, 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(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, 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(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5955e-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(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-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(2.9340e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2904e-05, device='cuda:0', grad_fn=<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(3.9301e-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(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(1.7895e-05, 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(0.0006, 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(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(2.3920e-05, device='cuda:0', grad_fn=<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(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-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(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7725e-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(0.0004, 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(1.8621e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-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(9.7467e-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(7.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3053e-05, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0155e-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(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-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(3.3366e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-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(1.8531e-05, 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(9.8592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3157e-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(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(2.4011e-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(1.9197e-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(9.0231e-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(1.1355e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1149e-05, device='cuda:0', grad_fn=<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(9.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9838e-05, device='cuda:0', grad_fn=<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(2.4556e-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(0.0002, 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(0.0006, 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(0.0002, 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(2.5434e-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(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2879e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0194e-05, 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(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(0.0001, 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(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-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(0.0004, 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(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0003, 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(0.0005, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0753e-05, device='cuda:0', grad_fn=<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(2.0408e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0004, 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(3.1126e-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(4.5810e-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(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-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(0.0001, 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(7.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5827e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0988e-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(0.0002, 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(3.1005e-05, device='cuda:0', grad_fn=<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(1.8470e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<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(0.0002, 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(8.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2843e-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(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-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(0.0001, 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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5971e-05, 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(0.0003, 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(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5404e-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(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1728e-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(1.3747e-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(1.1597e-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(0.0002, 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(8.2661e-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(2.9430e-05, 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(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(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(7.0530e-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(7.7383e-05, 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(0.0001, 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(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(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8481e-05, 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(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(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-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(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8209e-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(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(0.0005, 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(0.0006, 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(6.0830e-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(8.0201e-05, device='cuda:0', grad_fn=<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(1.7652e-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(1.2959e-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(1.4382e-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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-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(0.0006, 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(8.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1219e-05, device='cuda:0', grad_fn=<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(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7240e-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(8.8688e-05, 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(8.9928e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2422e-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(0.0003, 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(6.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-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(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(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, 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(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, 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(0.0001, 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(7.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6760e-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(0.0003, 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(0.0007, 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(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-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(7.1303e-05, device='cuda:0', grad_fn=<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(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9479, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(7.1934e-05, device='cuda:0', grad_fn=<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(0.8621, 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(2.8916e-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(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(9.4731e-05, device='cuda:0', grad_fn=<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(6.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-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(2.3769e-05, device='cuda:0', grad_fn=<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(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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7500e-05, 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(8.7882e-05, 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(0.0002, 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(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-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(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(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(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(5.0745e-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(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0926e-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(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-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(1.2566e-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(0.0001, 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(1.8591e-05, device='cuda:0', grad_fn=<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(5.2355e-05, device='cuda:0', grad_fn=<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(0.0003, 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(8.4427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<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(1.9318e-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(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, 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(0.0002, 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(1.3807e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-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(0.0001, 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(1.7138e-05, device='cuda:0', grad_fn=<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(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3802e-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(1.5836e-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(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7478e-05, device='cuda:0', grad_fn=<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(0.0008, 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(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7050e-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(2.2830e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(1.9681e-05, 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(1.4867e-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(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7705e-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(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, 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(2.2527e-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(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0651e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-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(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(5.9674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(9.1442e-06, 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(3.6122e-05, device='cuda:0', grad_fn=<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(2.9703e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-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(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, 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(7.9543e-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(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(7.1316e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2844e-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(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-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(7.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-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(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7649e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-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(7.1718e-05, 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(8.4252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5955e-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(1.9439e-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(9.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9628e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, 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(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, 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(0.0005, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-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(2.3254e-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(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(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_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(0.0009, 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(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7358e-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(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5089e-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(9.6011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4862e-05, device='cuda:0', grad_fn=<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(5.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-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(1.8924e-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(0.0005, 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(3.5213e-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(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(8.6490e-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(0.0002, 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(4.5533e-05, device='cuda:0', grad_fn=<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(2.9642e-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(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(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2091e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1260e-05, device='cuda:0', grad_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(0.0001, 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(8.3261e-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(3.5485e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5986e-05, device='cuda:0', grad_fn=<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(4.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9738e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5949e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-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(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(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, 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(0.0001, 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(7.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4341e-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(0.0005, 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(3.9066e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2609e-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(0.0003, 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(5.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1446e-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
				 {'precision': {'support': 0.84375, 'agreement': 0.6862745098039216, 'direct_attack': 0.35714285714285715, 'undercutter_attack': 0.5901639344262295, 'partial': 0.3076923076923077}, 'recall': {'support': 0.9869451697127938, 'agreement': 0.3125, 'direct_attack': 0.4166666666666667, 'undercutter_attack': 0.5538461538461539, 'partial': 0.23529411764705882}, 'f1': {'support': 0.9097472924187726, 'agreement': 0.4294478527607362, 'direct_attack': 0.3846153846153846, 'undercutter_attack': 0.5714285714285714, 'partial': 0.26666666666666666}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7429906542056075, 'recall': 0.7429906542056075, 'f1': 0.7429906542056075, 'support': None}, 'macro_avg': {'precision': 0.5570047218130633, 'recall': 0.5010504215745346, 'f1': 0.5123811535780263, 'support': None}, 'weighted_avg': {'precision': 0.7258318480376009, 'recall': 0.7429906542056075, 'f1': 0.7183838547922968, 'support': None}}
Loss: tensor(4.4053e-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(0.0001, 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(0.0001, 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(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(2.7886e-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(4.5598e-05, 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(1.4988e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, 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(4.0118e-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(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(1.0991e-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(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-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(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0006, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6666e-05, 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(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(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8886e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2863e-05, 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(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-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(0.0001, 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(0.0013, 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(0.0003, 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(7.8728e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5343e-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(7.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5268e-05, device='cuda:0', grad_fn=<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(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7067e-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(7.4883e-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(2.6130e-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(0.0002, device='cuda:0', grad_fn=<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(2.7008e-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(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9966e-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(6.9395e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-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(7.0068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6961e-05, device='cuda:0', grad_fn=<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(3.0490e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7603e-05, device='cuda:0', grad_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(9.3488e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5669e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, 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(9.5378e-06, 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(0.0002, 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(5.7825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4360e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, 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(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<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(6.6956e-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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-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(0.0004, 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(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3342e-05, device='cuda:0', grad_fn=<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(0.0007, 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(9.2916e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1695e-05, device='cuda:0', grad_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(0.0004, 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(7.5025e-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(0.0002, 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(0.0002, 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(2.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_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(7.6606e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, 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(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3104e-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(3.5939e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3208e-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(6.1957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2751e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<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(9.9920e-06, 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(0.0001, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4879e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8235e-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(5.5157e-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(5.1490e-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(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9300e-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(0.0002, 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(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-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(9.3593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(6.9882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<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(9.6963e-05, device='cuda:0', grad_fn=<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(2.1770e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0310e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-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(0.0002, 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(6.4645e-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(0.0003, 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(1.9409e-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(7.3907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<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(6.5913e-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(3.4244e-05, device='cuda:0', grad_fn=<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(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3513e-05, device='cuda:0', grad_fn=<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(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8272e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1491e-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(1.6229e-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(2.9370e-05, device='cuda:0', grad_fn=<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(8.3363e-05, 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(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-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(2.5827e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-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(1.1900e-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(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-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(0.0004, 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(2.5010e-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(0.0002, 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(1.2477, 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(5.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7068e-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(0.0003, 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(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1490e-05, 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(0.0001, 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(9.2750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6076e-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(3.6212e-05, 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(0.0004, 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(1.8833e-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(0.0001, 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(0.0007, 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(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6632e-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(0.0003, 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(8.4175e-06, 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(0.0008, 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(0.0009, 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(0.0004, 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(0.0001, 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(0.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3117e-05, 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(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, 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(0.0001, 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(3.2821e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-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(1.2959e-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(1.3323e-05, device='cuda:0', grad_fn=<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(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4005e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9324e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1836e-05, 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(0.0005, 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(1.1869e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<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(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-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(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6290e-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(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1762e-05, device='cuda:0', grad_fn=<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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4041e-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(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-05, device='cuda:0', grad_fn=<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(8.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, 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(0.0004, 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(0.0002, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-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(8.1771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1355e-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(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(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-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(0.0002, 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(8.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3055e-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(8.9380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-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(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-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(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6691e-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(1.7864e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7699e-05, device='cuda:0', grad_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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8675e-05, device='cuda:0', grad_fn=<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(2.7795e-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(0.0005, 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(8.1379e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0805e-05, 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(2.2406e-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(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-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(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-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(3.1489e-05, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, 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(0.0004, 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(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(0.0002, 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(5.0563e-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(2.2891e-05, 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(0.0004, 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(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<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(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-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(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7122e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-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(8.8345e-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(9.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6142e-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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4023e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0001, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-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(2.0771e-05, device='cuda:0', grad_fn=<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(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-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(0.0018, 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(1.5927e-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(8.1389e-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(4.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7522e-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(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4546e-05, 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(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(9.0158e-05, 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(7.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6065e-05, 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(5.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1442e-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(1.5291e-05, device='cuda:0', grad_fn=<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(1.0113e-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(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3024e-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(4.4686e-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(3.9239e-05, 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>)
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(6.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2562e-05, device='cuda:0', grad_fn=<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(7.4126e-05, device='cuda:0', grad_fn=<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(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, 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(8.5315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-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(6.7624e-05, 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(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0347e-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(7.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-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(0.0004, 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(3.3396e-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(1.2263e-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(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8261e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-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(1.0325e-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(7.9992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1748e-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(0.0001, 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(4.5957e-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(1.4625e-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(6.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5063e-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(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9776e-05, 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(7.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6591e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4510e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6257e-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(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(4.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4600e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4742e-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(2.8522e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8459821428571429, 'agreement': 0.7083333333333334, 'direct_attack': 0.3448275862068966, 'undercutter_attack': 0.5833333333333334, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.30357142857142855, 'direct_attack': 0.4166666666666667, 'undercutter_attack': 0.5384615384615384, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9121540312876054, 'agreement': 0.425, 'direct_attack': 0.37735849056603776, 'undercutter_attack': 0.5599999999999999, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7398753894080997, 'recall': 0.7398753894080997, 'f1': 0.7398753894080997, 'support': None}, 'macro_avg': {'precision': 0.5464952791461413, 'recall': 0.49082762448220907, 'f1': 0.5000637946933092, 'support': None}, 'weighted_avg': {'precision': 0.7263440574022068, 'recall': 0.7398753894080997, 'f1': 0.7151800948367, 'support': None}}
Loss: tensor(1.2687e-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(6.1640e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4127e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4043e-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(1.6199e-05, 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(1.4110e-05, 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(2.4071e-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(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(0.0004, 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(6.0314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, 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(0.0004, 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(2.3708e-05, device='cuda:0', grad_fn=<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(1.1294e-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(1.4382e-05, 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(4.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6677e-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(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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-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(1.3625e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-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(5.3598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-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(1.5987e-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(2.1589e-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(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(0.0003, 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(9.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4744e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9090e-05, 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(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-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(1.1839e-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(0.0001, 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(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7170e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1210e-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(3.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, 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(0.0005, 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(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2848e-05, 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(8.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3311e-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(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, 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(0.0005, 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(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, 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(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, 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(0.0007, 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(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(7.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-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(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-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(5.6418e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2120e-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(4.5705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<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(0.0003, 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(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, 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(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(7.7933e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8289e-05, 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(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(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2301e-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(0.0004, 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(0.0001, 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(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5665e-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(0.0002, 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(0.0002, 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(1.9711e-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(6.5148e-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(8.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7398e-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(2.3345e-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(9.2226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, 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(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-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(0.0002, 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(1.9984e-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(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(3.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, 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(9.5681e-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(5.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1411e-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(8.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(5.8421e-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(0.0001, 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(1.9045e-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(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-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(2.1649e-05, device='cuda:0', grad_fn=<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(2.5979e-05, device='cuda:0', grad_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(4.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0028e-05, device='cuda:0', grad_fn=<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(7.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4864e-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(1.9620e-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(1.0143e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0859e-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(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, 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(4.8753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8707e-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(5.2536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1616e-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(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(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9837e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, 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(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<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(9.6892e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<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(3.4514e-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(6.6483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4525e-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(5.5163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-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(8.3708e-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(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(0.0002, 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(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(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5523e-05, 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(2.1316e-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(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5814e-05, 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(0.0002, 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(0.0002, 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(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-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(0.0002, 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(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-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(8.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9887e-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(5.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9468e-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(0.0004, 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(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7767e-05, device='cuda:0', grad_fn=<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(2.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, 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(1.4685e-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(1.9318e-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(7.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(6.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7415e-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(7.0003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8060e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7285e-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(5.8632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, 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(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(2.1800e-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(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(0.0002, 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(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2474e-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(3.4516e-05, 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(6.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5563e-05, device='cuda:0', grad_fn=<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(8.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, 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(1.0931e-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(3.9179e-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(0.0002, 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(5.0684e-05, 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(5.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5608e-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(0.0002, 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(6.4715e-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(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(2.8250e-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(4.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-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(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-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(0.0004, 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(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-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(0.0002, 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(6.0257e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6571e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9614e-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(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4384e-05, device='cuda:0', grad_fn=<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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6342e-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(1.1324e-05, 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(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7882e-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(4.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2400e-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(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(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(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1543e-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(1.8591e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, 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(1.8197e-05, 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(0.0002, 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(7.0462e-05, device='cuda:0', grad_fn=<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(9.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-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(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, 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(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-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(6.6917e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9240e-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(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6613e-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(2.9679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-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(2.0650e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8728e-05, device='cuda:0', grad_fn=<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(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4831e-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(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-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(2.2709e-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(0.0002, 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(6.2750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3680e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5568e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1031e-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(5.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4669e-05, device='cuda:0', grad_fn=<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(4.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2932e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8694e-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(5.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1179e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4327e-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(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(9.8406e-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(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(6.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3455e-05, 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(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2349e-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(1.3898e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-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(3.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7100e-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(0.0002, 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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-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(1.7350e-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(5.2720e-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(8.8717e-06, 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(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-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(7.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2398e-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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1510e-05, device='cuda:0', grad_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(6.8430e-06, 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(0.0001, 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(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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5348e-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(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8361e-05, 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(0.0002, 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(0.0121, device='cuda:0', grad_fn=<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(8.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5065e-05, device='cuda:0', grad_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(3.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-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(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, 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(9.3737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7993e-05, device='cuda:0', grad_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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2127e-05, device='cuda:0', grad_fn=<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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0490e-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(7.4364e-05, 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(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7388e-05, device='cuda:0', grad_fn=<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(9.4470e-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(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0422e-05, device='cuda:0', grad_fn=<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(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8853e-05, 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(1.2566e-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(0.0004, 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(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-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(9.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-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(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6426e-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(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-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(8.4335e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-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(1.7138e-05, 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
				 {'precision': {'support': 0.8459821428571429, 'agreement': 0.7083333333333334, 'direct_attack': 0.32142857142857145, 'undercutter_attack': 0.5645161290322581, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.30357142857142855, 'direct_attack': 0.375, 'undercutter_attack': 0.5384615384615384, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9121540312876054, 'agreement': 0.425, 'direct_attack': 0.3461538461538462, 'undercutter_attack': 0.5511811023622047, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7367601246105919, 'recall': 0.7367601246105919, 'f1': 0.7367601246105919, 'support': None}, 'macro_avg': {'precision': 0.5380520353302611, 'recall': 0.48249429114887576, 'f1': 0.49205908628331185, 'support': None}, 'weighted_avg': {'precision': 0.7226894296935938, 'recall': 0.7367601246105919, 'f1': 0.7119541582662299, 'support': None}}
Loss: tensor(5.9600e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9645e-05, 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(1.7319e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9727e-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(9.7193e-05, 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(9.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3468e-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(8.8564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0508e-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(9.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5271e-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(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8956e-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(0.0002, 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(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8537e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2370e-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(6.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1806e-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(6.5968e-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(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9900e-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(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5121e-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(7.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1451e-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(9.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7756e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<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(7.6663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7216e-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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<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(5.9388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, 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(1.2657e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6090e-05, 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(3.6466e-05, 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(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6156e-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(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0439e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-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(9.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0004, 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(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2213e-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(0.0002, 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(0.0002, 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(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<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(1.4897e-05, device='cuda:0', grad_fn=<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(2.2406e-05, 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(0.0002, 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(9.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0511e-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(3.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5429e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0545e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2720e-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(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6938e-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(9.5091e-05, 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(9.3234e-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(0.0001, 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(1.0628e-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(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9962e-05, device='cuda:0', grad_fn=<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(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1272e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, 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(0.0001, 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(8.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9289e-05, device='cuda:0', grad_fn=<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(6.5306e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6767e-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(6.7851e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5576e-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(7.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1588e-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(4.1207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7245e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8198e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4179e-05, device='cuda:0', grad_fn=<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(7.7057e-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(9.7900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-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(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-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(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1027e-05, device='cuda:0', grad_fn=<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(6.0236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9909e-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(0.0003, 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(0.0002, 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(0.0003, 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(9.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2168e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8209e-05, device='cuda:0', grad_fn=<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(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-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(3.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-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(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9849e-05, 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(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-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(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3774e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1793e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7157e-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(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1588e-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(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9972e-05, device='cuda:0', grad_fn=<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(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3646e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1104e-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(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1870e-05, device='cuda:0', grad_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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-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(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-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(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6771e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5888e-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(4.7673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9039e-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(7.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0220e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1052e-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(0.0002, 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(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6331e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8153e-05, device='cuda:0', grad_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(1.3686e-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(1.5018e-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(1.1748e-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(0.0003, 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(3.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4801e-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(5.4566e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-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(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4558e-05, 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(1.4473e-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(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2688e-05, device='cuda:0', grad_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(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-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(6.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7187e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8974e-05, 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(5.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6829e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6977e-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(8.7663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-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(1.3989e-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(2.6948e-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(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(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1285e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4600e-05, device='cuda:0', grad_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(0.0002, 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(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-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(9.8400e-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(7.8596e-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(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1451e-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(1.0961e-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(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9920e-05, device='cuda:0', grad_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(1.0476e-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(1.2323e-05, 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(3.7332e-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(3.0071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3560e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9245e-05, 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(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(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2935e-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(2.3677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2175e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4325e-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(5.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7162e-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(0.0002, 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(9.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0369e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8434e-05, 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(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-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(7.7336e-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(7.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4229e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3148e-05, device='cuda:0', grad_fn=<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(9.3074e-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(8.4947e-05, 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(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9018e-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(7.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6909e-05, device='cuda:0', grad_fn=<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(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-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(6.9828e-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(2.2406e-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(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(4.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6222e-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(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7984e-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(8.6424e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(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(3.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7032e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3609e-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(7.7211e-06, 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(8.1197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-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(1.5139e-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(0.0008, 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(4.4001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-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(7.8152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8197e-05, 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(1.0022e-05, device='cuda:0', grad_fn=<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(8.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8662e-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(4.4273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7388e-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(2.8643e-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(3.6798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5141e-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(4.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2245e-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(5.1181e-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(5.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2799e-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(0.0005, 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(9.2653e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6823e-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(9.6590e-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(0.0002, 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(9.9012e-06, 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(5.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, 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(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(1.6078e-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(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4575e-05, 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(2.4586e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9596e-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(0.0002, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2724e-05, device='cuda:0', grad_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(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9924e-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(3.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2674e-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(3.0248e-05, device='cuda:0', grad_fn=<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(6.6397e-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(8.4340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, 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(3.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-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(0.0002, 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(5.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4186e-05, 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(1.0204e-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(8.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3661e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<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(3.3023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8087e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3794e-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(2.8613e-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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.32075471698113206, 'undercutter_attack': 0.5538461538461539, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5538461538461539, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.33663366336633666, 'undercutter_attack': 0.5538461538461539, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.735202492211838, 'recall': 0.735202492211838, 'f1': 0.735202492211838, 'support': None}, 'macro_avg': {'precision': 0.5341653051894821, 'recall': 0.47961883327341787, 'f1': 0.4884876601511454, 'support': None}, 'weighted_avg': {'precision': 0.7193521130551735, 'recall': 0.735202492211838, 'f1': 0.7091300582367677, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3169e-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(0.0002, 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(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-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(9.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5900e-05, device='cuda:0', grad_fn=<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(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9796e-05, device='cuda:0', grad_fn=<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(1.1385e-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(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5714e-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(7.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4835e-05, device='cuda:0', grad_fn=<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(8.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, 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(0.0001, 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(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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0522e-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(0.0004, 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(0.0002, 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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6128e-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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8012e-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(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_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(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-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(3.7787e-05, 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(0.0002, 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(2.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4287e-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(5.9541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-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(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(5.8118e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9917e-05, 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(0.0005, 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(6.7083e-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(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4620e-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(8.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7764e-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(9.4167e-06, 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(9.7801e-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(7.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4039e-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(1.0688e-05, 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(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, 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(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(8.4738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-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(1.2354e-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(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4424e-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(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9170e-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(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8467e-05, device='cuda:0', grad_fn=<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(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3855e-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(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0362e-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(2.5357e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5247e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-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(5.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6113e-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(8.9962e-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(0.0003, 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(7.9443e-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(3.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2797e-05, 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(0.0001, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6556e-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(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5187e-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(1.3625e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, 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(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7937e-05, 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(0.0002, 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(4.3541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5279e-05, 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(9.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, 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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-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(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1674e-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(0.0001, 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(5.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0483e-05, device='cuda:0', grad_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(4.7594e-05, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9109e-05, 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(3.3241e-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(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0797e-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(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8045e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0940e-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(0.0002, 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(9.4062e-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(2.7916e-05, 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(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2866e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4548e-05, 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(2.5706e-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(0.0001, 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(7.6908e-06, 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(0.0003, 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(9.5407e-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(0.0003, 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(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-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(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(6.3973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-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>)
Loss: tensor(1.0265e-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(7.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6304e-05, device='cuda:0', grad_fn=<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(4.0092e-05, 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(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1802e-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(9.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3599e-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(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0009, 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(0.0002, 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(7.0247e-06, 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(5.5891e-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(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9503e-05, device='cuda:0', grad_fn=<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(8.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-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(0.0004, 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(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6071e-05, device='cuda:0', grad_fn=<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(5.9347e-06, 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(7.3877e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9986e-05, device='cuda:0', grad_fn=<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(5.9217e-05, device='cuda:0', grad_fn=<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(6.4885e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2112e-05, device='cuda:0', grad_fn=<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(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8310e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-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(8.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-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(0.0003, 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(8.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3910e-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(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6405e-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(6.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0087e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9911e-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(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3803e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0261e-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(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, 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(3.1816e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2563e-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(2.0044e-05, 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(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0504e-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(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6350e-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(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5601e-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(0.0003, 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(0.0001, 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(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, 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(4.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0335e-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(0.0002, 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(8.5689e-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(5.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(6.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-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(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, 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(7.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4416e-05, 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(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8194e-05, 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(8.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0176e-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(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8780e-05, device='cuda:0', grad_fn=<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(8.7866e-05, device='cuda:0', grad_fn=<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(1.4049e-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(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0569e-05, device='cuda:0', grad_fn=<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(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3600e-05, 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(9.7113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, 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(6.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7094e-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(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(6.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-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(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4827e-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(8.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5529e-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(3.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-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(7.9229e-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(1.3958e-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(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(6.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7015e-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(2.3254e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6223e-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(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7179e-05, device='cuda:0', grad_fn=<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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5594e-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(5.9044e-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(8.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7164e-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(8.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, 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(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9499e-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(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8008e-05, 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(3.1024e-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(5.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8192e-05, device='cuda:0', grad_fn=<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(2.8834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5100e-05, device='cuda:0', grad_fn=<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(8.9427e-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(6.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3146e-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(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2983e-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(5.1103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, 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(3.7875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4877e-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(0.0001, 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(9.9812e-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(8.6017e-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(7.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-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(7.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, 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(0.0003, 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(9.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-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(9.3390e-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(6.2585e-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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4694e-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(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-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(3.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2622e-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(6.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, 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(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-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(6.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7213e-05, 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(1.8985e-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(6.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9583e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5717e-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(8.9578e-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(7.2053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7507e-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(7.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.3269230769230769, 'undercutter_attack': 0.5606060606060606, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.34, 'undercutter_attack': 0.564885496183206, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7367601246105919, 'recall': 0.7367601246105919, 'f1': 0.7367601246105919, 'support': None}, 'macro_avg': {'precision': 0.5367509585298524, 'recall': 0.48269575635034095, 'f1': 0.49136879594528854, 'support': None}, 'weighted_avg': {'precision': 0.7204977130810415, 'recall': 0.7367601246105919, 'f1': 0.7104994373805749, 'support': None}}
Loss: tensor(9.7428e-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(9.1745e-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(1.2051e-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(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8829e-05, device='cuda:0', grad_fn=<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(8.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0242e-05, device='cuda:0', grad_fn=<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(4.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-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(5.9069e-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(6.2430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1890e-05, device='cuda:0', grad_fn=<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(1.0234e-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(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6264e-05, device='cuda:0', grad_fn=<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(9.8147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3018e-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(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2126e-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(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5234e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6880e-05, 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(6.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8245e-05, 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(4.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9841e-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(9.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4767e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5901e-05, device='cuda:0', grad_fn=<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(5.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5876e-05, device='cuda:0', grad_fn=<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(3.6916e-05, device='cuda:0', grad_fn=<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(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5272e-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(7.6362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6226e-05, 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(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8274e-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(0.0001, device='cuda:0', grad_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(3.4203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8097e-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(4.4427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6946e-05, 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(0.0002, 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(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, 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(0.0001, 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(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2865e-05, device='cuda:0', grad_fn=<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(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-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(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1411e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4103e-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(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, 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(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, 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(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0858e-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(5.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(3.3182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7539e-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(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2349e-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(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1449e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(2.4249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-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(6.8255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6432e-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(5.1371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-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(9.7395e-05, 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(9.6481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-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(2.2137e-05, 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(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(7.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8284e-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(7.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, 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(5.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, 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(5.2637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2007e-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(1.7985e-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(6.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4614e-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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6930e-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(5.7848e-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(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5670e-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(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6578e-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(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, 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(6.9323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1103e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5890e-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(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2691e-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(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, 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(3.6434e-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(4.7107e-05, device='cuda:0', grad_fn=<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(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8276e-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(3.7662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-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(2.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-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(8.3463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3617e-05, device='cuda:0', grad_fn=<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(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0297e-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(8.4175e-06, 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(6.4380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4749e-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(1.3625e-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(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-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(5.5179e-05, 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(2.6100e-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(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.5412e-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(5.9507e-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(7.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1153e-05, device='cuda:0', grad_fn=<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(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, 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(1.1839e-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(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(8.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4735e-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(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7144e-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(8.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4332e-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(5.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8825e-05, 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(4.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4873e-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(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(4.7703e-05, 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(1.8894e-05, 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(4.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7622e-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(7.3968e-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(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(7.7817e-06, 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(6.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9467e-05, device='cuda:0', grad_fn=<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(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4418e-05, 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(7.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9450e-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(1.3565e-05, device='cuda:0', grad_fn=<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(7.8422e-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(3.4711e-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(7.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2766e-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(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-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(0.0001, 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(7.3524e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(8.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4712e-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(9.0849e-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(4.6202e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1247e-05, 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(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-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(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6999e-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(7.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2883e-05, device='cuda:0', grad_fn=<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(3.2589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4529e-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(0.0001, 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(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-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(4.0633e-05, 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(4.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1540e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0539e-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(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7400e-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(0.0004, 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(5.3400e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3520e-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(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8624e-05, 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(5.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6828e-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(5.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5882e-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(9.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8521e-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(7.8557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9347e-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(8.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0209e-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(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5723e-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(3.8362e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-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(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5779e-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(3.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-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(2.3796e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0565e-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(0.0003, 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(9.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-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(1.1809e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7666e-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(7.4654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-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(2.8551e-05, device='cuda:0', grad_fn=<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(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9559e-05, device='cuda:0', grad_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(4.7784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0132e-05, 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(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-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(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1575e-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(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3758e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7192e-05, 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(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2063e-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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6348e-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(6.6916e-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(9.4604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9907e-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(7.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0562e-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(7.9634e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9069e-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(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8140e-05, 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(3.5874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9068e-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(3.8850e-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(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1403e-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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.3269230769230769, 'undercutter_attack': 0.5606060606060606, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.34, 'undercutter_attack': 0.564885496183206, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7367601246105919, 'recall': 0.7367601246105919, 'f1': 0.7367601246105919, 'support': None}, 'macro_avg': {'precision': 0.5367509585298524, 'recall': 0.48269575635034095, 'f1': 0.49136879594528854, 'support': None}, 'weighted_avg': {'precision': 0.7204977130810415, 'recall': 0.7367601246105919, 'f1': 0.7104994373805749, 'support': None}}
Loss: tensor(6.6784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8414e-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(6.6289e-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(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5212e-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(2.0695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-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(8.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1321e-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(6.5046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-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(6.1882e-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(0.0003, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4828e-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(3.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4974e-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(1.1476e-05, 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(6.3104e-05, 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(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2923e-05, device='cuda:0', grad_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(0.0001, 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(0.0003, 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(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-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(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4900e-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(2.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9200e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-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(7.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, 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(2.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-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(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1752e-05, 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(7.1504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4277e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8823e-05, 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(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1023e-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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349e-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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-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(6.3829e-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(6.6426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0930e-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(6.9915e-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(5.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(9.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6629e-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(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2449e-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(7.8688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2378e-05, 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(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8104e-05, 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(2.1531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2934e-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(2.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-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(5.7898e-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(3.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7789e-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(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-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(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2614e-05, device='cuda:0', grad_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(1.1355e-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(3.3631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8339e-05, device='cuda:0', grad_fn=<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(2.8121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-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(5.3330e-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(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(5.7984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9166e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8429e-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(7.0385e-05, device='cuda:0', grad_fn=<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(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, 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(6.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9501e-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(7.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-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(9.6781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7497e-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(8.8953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6764e-05, 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(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, 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(7.7514e-06, 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(9.6590e-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(8.7809e-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(1.6956e-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(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1870e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0014e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0022e-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(9.7870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, 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(4.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1888e-05, 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(0.0003, 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(3.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, 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(4.0236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-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(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, 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(9.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9081e-05, device='cuda:0', grad_fn=<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(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6599e-05, device='cuda:0', grad_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(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6925e-05, 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(9.4007e-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(2.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, 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(8.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7885e-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(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8518e-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(1.7895e-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(7.1156e-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(4.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6102e-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(0.0001, 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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1302e-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(8.8411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8517e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2924e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2118e-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(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4079e-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(2.5438e-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(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4007e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-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(2.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9268e-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(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2352e-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(6.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8760e-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(6.6008e-06, 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(2.0589e-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(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7243e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9562e-05, 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(8.5665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, 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(3.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8806e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8761e-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(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2397e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2242e-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(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, 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(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2039e-05, 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(2.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-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(3.6874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-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(1.1476e-05, 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(4.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-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(4.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9874e-05, device='cuda:0', grad_fn=<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(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2584e-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(2.2224e-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(2.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1821e-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(6.6916e-06, 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(5.7624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9797e-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(1.3323e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3418e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0554e-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(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5315e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-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(4.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7403e-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(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3262e-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(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-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(5.1931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6655e-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(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, 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(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8090e-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(4.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2303e-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(5.0749e-05, 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(5.6420e-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(8.3041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-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(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7842e-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(2.1379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1408e-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(1.2959e-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(0.0002, 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(7.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6493e-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(0.0002, 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(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4764e-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(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6318e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, 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(6.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2046e-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(3.1516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8644e-05, 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(3.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5648e-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(6.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-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(4.3905e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, 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(2.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5910e-05, 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(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-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(3.8770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5322e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7514e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, 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(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4833e-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(4.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, 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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-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(9.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, 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(4.3705e-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(7.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3460e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8849e-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(7.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2328e-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(9.0534e-06, 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(2.5776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.3269230769230769, 'undercutter_attack': 0.5606060606060606, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.34, 'undercutter_attack': 0.564885496183206, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7367601246105919, 'recall': 0.7367601246105919, 'f1': 0.7367601246105919, 'support': None}, 'macro_avg': {'precision': 0.5367509585298524, 'recall': 0.48269575635034095, 'f1': 0.49136879594528854, 'support': None}, 'weighted_avg': {'precision': 0.7204977130810415, 'recall': 0.7367601246105919, 'f1': 0.7104994373805749, 'support': None}}
Loss: tensor(4.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, 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(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-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(4.8349e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9935e-05, device='cuda:0', grad_fn=<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(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3133e-05, 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(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5935e-05, 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(6.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6765e-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(4.1510e-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(2.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6500e-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(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4935e-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(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8378e-05, device='cuda:0', grad_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(3.2537e-05, device='cuda:0', grad_fn=<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(0.0001, 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(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3432e-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(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5936e-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(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-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(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6988e-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(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-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(8.4211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3124e-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(7.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-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(0.0002, 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(7.6120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8997e-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(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1962e-05, 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(7.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1347e-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(3.2251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1846e-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(6.0388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1415e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1508e-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(7.7693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0704e-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(3.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-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(9.4167e-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(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4715e-05, 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(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8227e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392e-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(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-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(7.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-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(4.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5257e-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(3.7641e-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(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, 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(7.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5217e-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(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0326e-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(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2632e-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(4.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, 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(0.0001, 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(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, 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(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-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(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1091e-05, device='cuda:0', grad_fn=<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(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-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(6.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-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(6.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0082e-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(1.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7892e-05, 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(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-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(5.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4458e-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(5.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4903e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, 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(1.7524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4250e-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(1.4655e-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(4.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, 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(1.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5761e-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(2.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9677e-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(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7551e-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(2.6892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, 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(6.8602e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9304e-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(6.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6221e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5348e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9928e-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(2.4344e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0698e-05, 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(1.0204e-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(7.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9515e-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>)
Loss: tensor(2.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0328e-05, 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(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, 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(6.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2360e-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(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-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(8.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9078e-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(4.5078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4488e-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(3.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6856e-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(3.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0612e-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(3.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5241e-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(4.2415e-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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3690e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1857e-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(5.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5077e-05, device='cuda:0', grad_fn=<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(1.3959e-05, 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(7.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-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(4.3821e-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(4.4425e-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(6.1023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8819e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4061e-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(0.0001, 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(8.9084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1632e-05, device='cuda:0', grad_fn=<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(6.5246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2043e-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(7.3538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1457e-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(7.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, 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(2.9175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6063e-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(4.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5301e-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(8.0076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3269e-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(7.8973e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, 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(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, 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(2.9985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4049e-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(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0912e-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(9.0866e-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(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8032e-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(8.5626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5038e-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(6.4158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2079e-05, 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(6.1984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3795e-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(4.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-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(9.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1298e-05, device='cuda:0', grad_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(5.0263e-06, 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(8.1615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0562e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0086e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2671e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, 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(6.8978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-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(4.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2053e-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(9.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, 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(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8147e-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(7.2367e-06, 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(7.7328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3410e-05, 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(6.2980e-06, 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(5.2142e-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(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, 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(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0894e-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(1.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, 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(6.3001e-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(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-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(3.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7590e-05, 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(7.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-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(2.4232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8966e-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(1.4140e-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(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2250e-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(9.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0803e-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(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, 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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.3269230769230769, 'undercutter_attack': 0.5606060606060606, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.34, 'undercutter_attack': 0.564885496183206, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7367601246105919, 'recall': 0.7367601246105919, 'f1': 0.7367601246105919, 'support': None}, 'macro_avg': {'precision': 0.5367509585298524, 'recall': 0.48269575635034095, 'f1': 0.49136879594528854, 'support': None}, 'weighted_avg': {'precision': 0.7204977130810415, 'recall': 0.7367601246105919, 'f1': 0.7104994373805749, 'support': None}}
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, 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(4.4073e-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(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-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(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7625e-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(4.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2020e-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(0.0001, 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(5.4191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7574e-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(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-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(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, 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(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1036e-05, 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(3.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1089e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2323e-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(9.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4153e-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(5.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4609e-05, 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(6.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-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(3.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0351e-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(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9217e-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(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-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(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7173e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5224e-05, 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(3.3255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1267e-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(2.1144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_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(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6997e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6143e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2398e-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(6.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1052e-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(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5711e-05, 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(5.2011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2928e-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(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-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(2.8519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9310e-05, device='cuda:0', grad_fn=<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(6.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, 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(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8738e-05, 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(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-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(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1537e-05, 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(9.9273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3034e-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(4.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3286e-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(8.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9714e-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(3.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9289e-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(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0033e-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(9.1139e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0097e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7542e-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(4.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8870e-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(8.4781e-06, 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(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, 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(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9271e-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(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9100e-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(3.8432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6400e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, 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(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-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(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7669e-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(7.3992e-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(3.9670e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4811e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7566e-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(1.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-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(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, 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(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, 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(3.8757e-06, 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(2.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6634e-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(5.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0901e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3858e-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(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6040e-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(3.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5042e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2691e-05, 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(7.1637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5645e-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(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7456e-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(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0140e-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(9.1442e-06, 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(7.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5123e-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(6.2794e-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(8.0393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1032e-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(7.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2691e-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(5.1720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-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(3.6918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6629e-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(6.4502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8998e-05, 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(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(2.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9082e-05, device='cuda:0', grad_fn=<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(7.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6170e-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(5.7967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9799e-05, 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(4.7274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4447e-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(3.9363e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-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(5.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6248e-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(5.4059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9513e-05, 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(2.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4552e-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(1.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5606e-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(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0411e-05, 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(9.6269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-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(5.9254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7027e-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(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6187e-05, 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(9.0932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7888e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1666e-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(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3141e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7558e-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(8.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, 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(4.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8426e-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(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3197e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4085e-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(1.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1719e-05, 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(6.4042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8597e-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(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-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(6.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, 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(1.5653e-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(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6391e-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(6.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1515e-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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.3269230769230769, 'undercutter_attack': 0.5606060606060606, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5692307692307692, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.34, 'undercutter_attack': 0.564885496183206, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7367601246105919, 'recall': 0.7367601246105919, 'f1': 0.7367601246105919, 'support': None}, 'macro_avg': {'precision': 0.5367509585298524, 'recall': 0.48269575635034095, 'f1': 0.49136879594528854, 'support': None}, 'weighted_avg': {'precision': 0.7204977130810415, 'recall': 0.7367601246105919, 'f1': 0.7104994373805749, 'support': None}}
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6683e-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(5.2962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7035e-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(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5911e-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(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0739e-05, 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(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9667e-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(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2386e-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(4.8741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-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(3.7849e-06, 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(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, 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(5.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3716e-05, 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(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-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(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6700e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3912e-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(2.3287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8689e-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(1.8318e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-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(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3455e-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(3.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4710e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, 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(6.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6000e-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(6.8976e-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(2.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, 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(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1008e-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(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-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(1.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6492e-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(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5426e-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(7.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2095e-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(7.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0627e-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(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6906e-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(4.8883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4800e-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(6.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7189e-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(9.0116e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1284e-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(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1440e-05, 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(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8439e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8852e-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(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2971e-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(4.0850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6148e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4621e-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(4.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8359e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1580e-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(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3651e-05, 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(3.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-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(6.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6114e-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(2.9252e-05, 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(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7585e-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(3.6732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9222e-05, device='cuda:0', grad_fn=<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(7.0172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2928e-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(3.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3239e-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(2.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6524e-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(2.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1627e-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(8.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, 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(6.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6081e-05, device='cuda:0', grad_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(3.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, 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(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5154e-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(5.4751e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9243e-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(8.6603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3280e-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(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-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(5.2508e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1405e-05, 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(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9407e-05, 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(3.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-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(6.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2667e-05, 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(5.7530e-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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5227e-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(1.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8512e-05, 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(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6266e-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(3.6091e-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(4.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2627e-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(6.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1984e-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(3.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4018e-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(8.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8040e-05, 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(6.4797e-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(8.7421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-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(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-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(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8355e-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(1.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1226e-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(6.6331e-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(2.8856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-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(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7749e-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(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1114e-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(2.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, 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(6.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-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(5.8070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8832e-05, 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(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3910e-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(8.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2055e-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>)
			------------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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.7021276595744681, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.5671641791044776, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5846153846153846, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.4150943396226415, 'direct_attack': 0.3434343434343434, 'undercutter_attack': 0.5757575757575757, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7383177570093458, 'recall': 0.7383177570093458, 'f1': 0.7383177570093457, 'support': None}, 'macro_avg': {'precision': 0.5393446335115872, 'recall': 0.485772679427264, 'f1': 0.4942300805470312, 'support': None}, 'weighted_avg': {'precision': 0.7216409685484706, 'recall': 0.7383177570093458, 'f1': 0.7118569664416069, 'support': None}}
Loss: tensor(5.8205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7844e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1029e-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(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0997e-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(3.6408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0798e-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(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(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, 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(5.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.6746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0658e-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(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9060e-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(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8207e-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(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3517e-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(8.5990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1932e-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(4.2545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6071e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4470e-05, 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(9.0112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8023e-05, 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(2.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1585e-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(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5832e-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(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4983e-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(9.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0413e-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(2.0289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3002e-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(7.2266e-05, 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(6.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0756e-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(3.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9097e-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(7.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4673e-05, 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(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7148e-05, 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(5.3027e-05, 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(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6793e-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(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4066e-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(3.4981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8386e-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(5.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, 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(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5562e-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(4.0534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0700e-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(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9817e-05, 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(4.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1898e-05, 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(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-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(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3564e-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(2.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1748e-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(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6952e-05, 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(7.9034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, 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(2.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9003e-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(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3940e-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(4.5056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, 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(1.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6206e-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(3.5384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-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(2.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7360e-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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-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(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6522e-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(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4716e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5334e-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(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0732e-05, 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(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0217e-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(1.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7438e-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(9.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, 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(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, 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(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5517e-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(2.9371e-06, 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(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3972e-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(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1571e-05, 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(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1458e-05, device='cuda:0', grad_fn=<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(1.5429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3185e-05, 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(3.1274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7925e-05, 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(5.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7566e-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(1.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2540e-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(4.2196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-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(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8668e-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(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8507e-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(3.3346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2099e-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(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-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
				 {'precision': {'support': 0.844097995545657, 'agreement': 0.717391304347826, 'direct_attack': 0.3269230769230769, 'undercutter_attack': 0.5671641791044776, 'partial': 0.25}, 'recall': {'support': 0.9895561357702349, 'agreement': 0.29464285714285715, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.5846153846153846, 'partial': 0.20588235294117646}, 'f1': {'support': 0.9110576923076923, 'agreement': 0.41772151898734183, 'direct_attack': 0.34, 'undercutter_attack': 0.5757575757575757, 'partial': 0.22580645161290322}, 'support': {'support': 383, 'agreement': 112, 'direct_attack': 48, 'undercutter_attack': 65, 'partial': 34}, 'micro_avg': {'precision': 0.7383177570093458, 'recall': 0.7383177570093458, 'f1': 0.7383177570093457, 'support': None}, 'macro_avg': {'precision': 0.5411153111842075, 'recall': 0.485772679427264, 'f1': 0.49406864773310255, 'support': None}, 'weighted_avg': {'precision': 0.7238245135748315, 'recall': 0.7383177570093458, 'f1': 0.7120585172266504, '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
				 {'precision': {'support': 0.5859872611464968, 'agreement': 0.0851063829787234, 'direct_attack': 0.08597285067873303, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.34074074074074073, 'agreement': 0.044444444444444446, 'direct_attack': 0.4523809523809524, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.43091334894613587, 'agreement': 0.05839416058394161, 'direct_attack': 0.14448669201520914, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.25219298245614036, 'recall': 0.25219298245614036, 'f1': 0.25219298245614036, 'support': None}, 'macro_avg': {'precision': 0.15141329896079064, 'recall': 0.16751322751322753, 'f1': 0.12675884030905732, 'support': None}, 'weighted_avg': {'precision': 0.3716820059345308, 'recall': 0.25219298245614036, 'f1': 0.2799792099400224, 'support': None}}
Loss: tensor(1.9548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, 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(1.8771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7064, 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
				 {'precision': {'support': 0.5921052631578947, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 1.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.743801652892562, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.5921052631578947, 'recall': 0.5921052631578947, 'f1': 0.5921052631578947, 'support': None}, 'macro_avg': {'precision': 0.11842105263157894, 'recall': 0.2, 'f1': 0.1487603305785124, 'support': None}, 'weighted_avg': {'precision': 0.35058864265927975, 'recall': 0.5921052631578947, 'f1': 0.44040887342322743, 'support': None}}
Loss: tensor(1.9677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9021, 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
				 {'precision': {'support': 0.5921052631578947, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 1.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.743801652892562, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.5921052631578947, 'recall': 0.5921052631578947, 'f1': 0.5921052631578947, 'support': None}, 'macro_avg': {'precision': 0.11842105263157894, 'recall': 0.2, 'f1': 0.1487603305785124, 'support': None}, 'weighted_avg': {'precision': 0.35058864265927975, 'recall': 0.5921052631578947, 'f1': 0.44040887342322743, 'support': None}}
Loss: tensor(0.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4305, 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
				 {'precision': {'support': 0.659846547314578, 'agreement': 0.660377358490566, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.9555555555555556, 'agreement': 0.3888888888888889, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.7806354009077157, 'agreement': 0.4895104895104895, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.6425438596491229, 'recall': 0.6425438596491229, 'f1': 0.6425438596491229, 'support': None}, 'macro_avg': {'precision': 0.2640447811610288, 'recall': 0.2688888888888889, 'f1': 0.25402917808364106, 'support': None}, 'weighted_avg': {'precision': 0.5210362500857172, 'recall': 0.6425438596491229, 'f1': 0.5588322418882178, 'support': None}}
Loss: tensor(4.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, 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
				 {'precision': {'support': 0.917910447761194, 'agreement': 0.6987951807228916, 'direct_attack': 0.3088235294117647, 'undercutter_attack': 0.3125, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9111111111111111, 'agreement': 0.6444444444444445, 'direct_attack': 0.5, 'undercutter_attack': 0.1724137931034483, 'partial': 0.12}, 'f1': {'support': 0.9144981412639405, 'agreement': 0.6705202312138728, 'direct_attack': 0.38181818181818183, 'undercutter_attack': 0.22222222222222224, 'partial': 0.13043478260869565}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7302631578947368, 'recall': 0.7302631578947368, 'f1': 0.7302631578947368, 'support': None}, 'macro_avg': {'precision': 0.4761772601505986, 'recall': 0.4695938697318008, 'f1': 0.46389871182538267, 'support': None}, 'weighted_avg': {'precision': 0.7375699648405818, 'recall': 0.7302631578947368, 'f1': 0.7302697293783728, 'support': None}}
Loss: tensor(0.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1367, 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
				 {'precision': {'support': 0.9306569343065694, 'agreement': 0.7228915662650602, 'direct_attack': 0.25, 'undercutter_attack': 0.2916666666666667, 'partial': 0.2608695652173913}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.6666666666666666, 'direct_attack': 0.023809523809523808, 'undercutter_attack': 0.7241379310344828, 'partial': 0.24}, 'f1': {'support': 0.9375, 'agreement': 0.6936416184971098, 'direct_attack': 0.04347826086956521, 'undercutter_attack': 0.4158415841584159, 'partial': 0.24999999999999994}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7521929824561403, 'recall': 0.7521929824561403, 'f1': 0.7521929824561403, 'support': None}, 'macro_avg': {'precision': 0.49121694649113745, 'recall': 0.5198117131910236, 'f1': 0.46809229270501823, 'support': None}, 'weighted_avg': {'precision': 0.7496001879175379, 'recall': 0.7521929824561403, 'f1': 0.7361584178988063, 'support': None}}
Loss: tensor(0.8713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7251, 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(0.8808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3748, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, 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(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(11.2379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, 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
				 {'precision': {'support': 0.8668831168831169, 'agreement': 0.7777777777777778, 'direct_attack': 0.2, 'undercutter_attack': 0.2972972972972973, 'partial': 0.0}, 'recall': {'support': 0.9888888888888889, 'agreement': 0.4666666666666667, 'direct_attack': 0.09523809523809523, 'undercutter_attack': 0.7586206896551724, 'partial': 0.0}, 'f1': {'support': 0.9238754325259516, 'agreement': 0.5833333333333334, 'direct_attack': 0.12903225806451613, 'undercutter_attack': 0.42718446601941745, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7346491228070176, 'recall': 0.7346491228070176, 'f1': 0.7346491228070176, 'support': None}, 'macro_avg': {'precision': 0.42839163839163835, 'recall': 0.46188286808976464, 'f1': 0.4126850979886437, 'support': None}, 'weighted_avg': {'precision': 0.704122945570314, 'recall': 0.7346491228070176, 'f1': 0.7012150682791224, 'support': None}}
Loss: tensor(1.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, 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(2.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, 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(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7730, 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(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, 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(3.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, 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(1.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5100, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8317, 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(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, 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(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2832, 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(3.9421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2508, 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(0.8607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5605, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, 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(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3067, 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(0.7685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9090, 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(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6688, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5427, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.8867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3433, 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(0.1532, 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(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, 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(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4848, 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(0.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8721, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, 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(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, 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(2.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2815, 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(0.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, 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
				 {'precision': {'support': 0.8783783783783784, 'agreement': 0.6470588235294118, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.3333333333333333, 'partial': 0.0}, 'recall': {'support': 0.9629629629629629, 'agreement': 0.4888888888888889, 'direct_attack': 0.19047619047619047, 'undercutter_attack': 0.7241379310344828, 'partial': 0.0}, 'f1': {'support': 0.9187279151943463, 'agreement': 0.5569620253164557, 'direct_attack': 0.22857142857142854, 'undercutter_attack': 0.45652173913043476, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7302631578947368, 'recall': 0.7302631578947368, 'f1': 0.7302631578947368, 'support': None}, 'macro_avg': {'precision': 0.42889696419108186, 'recall': 0.473293194672505, 'f1': 0.432156621642533, 'support': None}, 'weighted_avg': {'precision': 0.6953160590931489, 'recall': 0.7302631578947368, 'f1': 0.7039961618766166, 'support': None}}
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, 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(0.5778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.4886, 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(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, 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(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2541, 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(1.5260, 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(0.0094, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, 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(2.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5609, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, 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(0.0046, 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(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, 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(1.6262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, 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(0.0831, 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(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, 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(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, 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(0.1153, 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(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, 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(0.0887, 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(0.0010, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1276, 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(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, 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(0.6564, 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(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4724, 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(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, 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(0.1294, 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(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, 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(0.0011, 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(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2932, 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(0.0672, 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(0.0017, 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(1.9806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3885, 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(0.0006, 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(0.0027, 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(1.0161, 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(0.0014, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9499, 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(0.0051, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2786, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4375, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2964, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, 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(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0053, 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(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3510, 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(0.4366, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, 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(0.0401, 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(0.2187, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, 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(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5389, 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(1.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, 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(3.0161, 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(0.0076, 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(0.5071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, 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(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0022, 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(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2368, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, 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(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604, 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(0.0026, 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(0.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, 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(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5982, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6259, 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(1.1889, 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(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, 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(0.0042, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1548, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5820, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4148, 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(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6842, 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(0.0042, 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(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(1.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6139, 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(0.0007, 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(5.6398, 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(5.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2755, 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(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, 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(1.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, 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(0.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3577, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, 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(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4677, 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(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(2.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, 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(0.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3890, 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(0.8598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, 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(1.4362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, 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(0.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9875, 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(1.2244, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3562, 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(1.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2116, 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(0.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, 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(1.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, 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(0.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5760, 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(0.0893, 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(2.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1996, 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(0.9423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(1.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9009, device='cuda:0', grad_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(0.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0090, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8463, 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(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, 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(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8854, 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(1.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8944, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, 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(0.0021, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4606, 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(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6500, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4745, 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(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, 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(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5316, 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(1.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5057, 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(0.0136, 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(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, 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(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, 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
				 {'precision': {'support': 0.9372693726937269, 'agreement': 0.6951219512195121, 'direct_attack': 0.34615384615384615, 'undercutter_attack': 0.32857142857142857, 'partial': 0.0}, 'recall': {'support': 0.9407407407407408, 'agreement': 0.6333333333333333, 'direct_attack': 0.21428571428571427, 'undercutter_attack': 0.7931034482758621, 'partial': 0.0}, 'f1': {'support': 0.9390018484288354, 'agreement': 0.6627906976744186, 'direct_attack': 0.2647058823529412, 'undercutter_attack': 0.46464646464646464, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7521929824561403, 'recall': 0.7521929824561403, 'f1': 0.7521929824561403, 'support': None}, 'macro_avg': {'precision': 0.4614233197277028, 'recall': 0.51629264732713, 'f1': 0.4662289786205319, 'support': None}, 'weighted_avg': {'precision': 0.7449358315879282, 'recall': 0.7521929824561403, 'f1': 0.7407325798246802, 'support': None}}
Loss: tensor(1.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0123, 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(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7461, 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(1.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4460, 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(0.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, 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(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, 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(0.4709, 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(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1618, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, 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(0.4732, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2922, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, 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(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, 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(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5952, 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(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0648, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3655, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, 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(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, 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(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, 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(0.1816, 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(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2324, 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(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, 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(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3494, 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(0.0047, 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(0.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4270, 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(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.9392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2162, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0071, 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(1.4139, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1850, 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(0.0196, 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(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, 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(4.1294, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9529, 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(0.0062, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1310, 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(3.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, 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(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0058, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4395, 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(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0708, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, 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(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6629, 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(0.0208, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0642, 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(0.0037, 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(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1857, 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(0.0013, 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(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, 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(4.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4292, 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(0.0193, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, 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(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, 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(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, 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(0.0492, 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(0.0028, 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(0.6015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, 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(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9396, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4346, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.1076, 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(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, 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(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, 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(0.0032, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851, 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(0.4470, 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(0.7767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8545, 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(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, 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(0.0131, 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(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0164, 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(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, 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(1.5031, 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(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, 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(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, 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(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4874, 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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1665, 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(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5937, device='cuda:0', grad_fn=<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(0.9249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, 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(0.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6604, 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(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, 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(1.0059, 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(0.5220, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, 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(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8162, device='cuda:0', grad_fn=<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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(1.8543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0173, 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(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4505, 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(0.0124, 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(0.0776, 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(0.0032, 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(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, 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(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4423, 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(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, 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(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2958, 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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, 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(0.0017, 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(0.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, 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(0.0063, 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(5.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, 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(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4605, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, 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(0.0512, 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(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, 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(0.0055, 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(0.0563, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4851, 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(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, 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(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, 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(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, 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(0.1917, 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(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, 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(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.1228, 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(0.0003, 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(0.0393, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1967, 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(0.0057, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, 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(0.0119, 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(0.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1401, 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(0.0049, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, 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(0.0011, 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(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8988, 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(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, 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(0.0010, 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(0.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5175, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, 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(0.0016, 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(9.8882e-05, 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(0.0106, 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(0.0007, 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(1.0212, 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(0.0005, 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(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8514, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, 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(0.0307, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(0.0005, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, 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(1.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8440, 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(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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
				 {'precision': {'support': 0.8673139158576052, 'agreement': 0.7272727272727273, 'direct_attack': 0.4117647058823529, 'undercutter_attack': 0.34782608695652173, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9925925925925926, 'agreement': 0.5333333333333333, 'direct_attack': 0.5, 'undercutter_attack': 0.27586206896551724, 'partial': 0.08}, 'f1': {'support': 0.92573402417962, 'agreement': 0.6153846153846153, 'direct_attack': 0.45161290322580644, 'undercutter_attack': 0.3076923076923077, 'partial': 0.125}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7609649122807017, 'recall': 0.7609649122807017, 'f1': 0.7609649122807016, 'support': None}, 'macro_avg': {'precision': 0.5279783443366985, 'recall': 0.4763575989782887, 'f1': 0.48508477009646994, 'support': None}, 'weighted_avg': {'precision': 0.732792179929285, 'recall': 0.7609649122807017, 'f1': 0.7376066244992843, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0010, 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(0.1052, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2914, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, 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(0.0051, 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(0.0189, 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(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, 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(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0171, 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(0.0002, 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(2.5285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, 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(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, 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(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<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(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<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(0.0005, 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(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0174, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, 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(0.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4419, 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(0.8816, 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(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8278, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3074, 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(0.0018, 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(0.0457, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7115e-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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, 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(0.0009, 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(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, 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(0.0004, 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(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8906, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3254, 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(0.5852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0025, 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(0.0019, 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(0.0178, 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(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, 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(1.9291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(7.9961e-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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, 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(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, 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(0.0006, 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(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3684, 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(0.0037, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3443, 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(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8853, 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(0.0075, 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(0.4583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, 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(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, 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(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, 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(0.0253, 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(0.0221, 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(0.0635, 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(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, 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(0.0224, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1180, 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(0.0067, 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(0.0027, 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(0.0002, 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(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, 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(0.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9933, 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(0.0035, device='cuda:0', grad_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(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0171, 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(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0808, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4942, 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(0.0076, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0034, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5904, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, 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(0.0062, 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(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4791, 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(0.0019, 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(0.0007, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, 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(0.0190, 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(0.0100, 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(1.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, 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(0.0021, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, 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(0.0005, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(0.3044, 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(0.0024, 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(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6296, 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(0.0131, 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(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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3839, 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(0.0046, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, 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(0.0053, 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(0.1315, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0785, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5322, 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(1.5211, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, 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(0.0496, 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(0.0233, 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(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, 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(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, 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(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2790, 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(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8884, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0274, 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(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, 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(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, 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(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, 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(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0162, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8850, 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(0.0046, 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(0.0027, 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>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, 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(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, 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(0.0521, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1603, 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(0.0162, 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(0.0124, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, 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(0.0146, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0060, 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(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.7690, 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(0.0108, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6732, 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(0.3638, 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(0.0034, 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(0.0602, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, 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(0.6414, 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(0.0051, 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(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, 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(6.3163, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198, 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(0.0002, 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(0.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2741, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, 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(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, 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(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, 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(0.2572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9141, 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(0.0036, 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(0.5006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3132, 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(3.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, 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(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4460, 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(0.0074, 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(0.0700, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2914, 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(1.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, 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(0.2725, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5482, 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(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.9198, 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(0.3061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8278, 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(0.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8792, 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(0.0052, 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(0.0063, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, 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(0.0213, 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(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, 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(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9878, 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(1.1067, 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(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7893e-05, 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(0.0361, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, 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(0.6022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9690, 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(0.0030, 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(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, 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(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2791, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, 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(0.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, 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(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(9.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_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(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4388, 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(2.1797, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, 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(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0360, 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(0.0122, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, 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(0.0190, device='cuda:0', grad_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(0.9449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, 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(0.2049, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, 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(0.0068, 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(1.8596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2762, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, 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
				 {'precision': {'support': 0.8737864077669902, 'agreement': 0.7543859649122807, 'direct_attack': 0.3877551020408163, 'undercutter_attack': 0.38235294117647056, 'partial': 0.2857142857142857}, 'recall': {'support': 1.0, 'agreement': 0.4777777777777778, 'direct_attack': 0.4523809523809524, 'undercutter_attack': 0.4482758620689655, 'partial': 0.08}, 'f1': {'support': 0.932642487046632, 'agreement': 0.5850340136054423, 'direct_attack': 0.41758241758241765, 'undercutter_attack': 0.41269841269841273, 'partial': 0.125}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7609649122807017, 'recall': 0.7609649122807017, 'f1': 0.7609649122807016, 'support': None}, 'macro_avg': {'precision': 0.5367989403221687, 'recall': 0.49168691844553913, 'f1': 0.49459146618658095, 'support': None}, 'weighted_avg': {'precision': 0.7419602492585124, 'recall': 0.7609649122807017, 'f1': 0.7392505443723596, 'support': None}}
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, 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(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, 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(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, 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(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, 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(0.0018, 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(0.0182, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, 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(0.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0175, 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(0.0015, 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(0.0072, 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(0.0047, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0168, 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(0.0020, 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(0.0006, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0424e-05, 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(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3210e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, 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(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.0015, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, 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(0.0381, 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(0.0113, 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(8.3775e-05, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2415, 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(0.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3109e-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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, 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(0.8864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4485, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(0.0081, 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(0.0003, 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(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, 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(0.0011, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8447e-05, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, 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(0.0073, 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(0.0027, 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(0.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, 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(0.0046, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0009, 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(0.0017, 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(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, 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(0.0030, 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(0.0015, 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(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, 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(0.0007, 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(0.0003, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, 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(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, 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(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<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(0.0119, 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(0.0154, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, 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(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, 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(0.0148, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3620, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.1807, 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(0.0105, 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(0.1315, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, 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(0.0514, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1404e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, 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(0.0036, 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(0.0149, 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(0.0007, 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(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4893, 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(0.0059, 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(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, 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(0.0842, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, 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(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1838e-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(9.2614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, 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(0.0017, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6459e-05, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0008, 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(0.0024, 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(0.0004, 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(0.0004, 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(0.0041, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, 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(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5824, 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(0.0054, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0704, 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(0.0011, 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(0.0043, 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(5.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, 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(0.0046, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1746e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5690, 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(0.0129, 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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, 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(0.0060, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5700, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0008, 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(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(0.0005, 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(0.0002, 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(0.0775, 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(0.9408, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, 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(0.0005, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, 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(0.0024, 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(0.0034, 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(0.0856, 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(0.0004, 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(9.2252e-05, 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(0.0004, 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(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0009, 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(0.0002, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, 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(0.0007, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, 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(0.0177, 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(0.0004, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1815e-05, 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(0.0334, 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(0.0474, 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(0.0029, 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(0.0056, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4069e-05, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.0023, 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(0.0034, 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(0.0005, 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(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0042, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.0003, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, 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(0.0053, 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(0.0039, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, 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(0.0187, 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(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1208, 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(0.0885, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, 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(9.9003e-05, 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(0.0096, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, 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(0.0332, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-05, 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(0.0033, 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(0.0002, 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(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3239e-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(0.0001, 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(0.0028, 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(9.5640e-05, 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(0.0095, 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(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5025e-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(0.0002, 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(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5914e-05, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9284, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, 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(0.0027, 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(0.0059, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-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(0.0007, 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(0.0297, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, 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(0.0136, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3500, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, 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(0.0022, 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(0.0004, 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(0.3333, 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(0.0295, device='cuda:0', grad_fn=<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(2.1952e-05, 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(0.1774, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9182e-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(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, 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(0.0049, 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(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9161e-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(0.0003, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, 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(0.0145, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-05, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0034, 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(4.2631e-05, 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(6.7395e-05, 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(2.6070e-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(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-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(0.0021, 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(0.0041, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, 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(0.0023, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-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(0.0024, 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(9.1615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, 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(0.0261, 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(0.0065, 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(0.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-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(6.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2007, 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(0.0014, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, 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(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-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(0.1545, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-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(6.1039e-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(0.0002, 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(0.0128, 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(0.0016, 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(9.7762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, 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(7.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, 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(5.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9576, 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(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-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(1.4581, 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(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-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(0.1130, 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(0.2239, 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(0.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2306, device='cuda:0', grad_fn=<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(0.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<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(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, 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(0.5998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, 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(0.0004, 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(1.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(1.0093, device='cuda:0', grad_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(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1858, 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(9.1889e-05, 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(0.0312, 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
				 {'precision': {'support': 0.8795986622073578, 'agreement': 0.7111111111111111, 'direct_attack': 0.3958333333333333, 'undercutter_attack': 0.35714285714285715, 'partial': 0.09090909090909091}, 'recall': {'support': 0.9740740740740741, 'agreement': 0.35555555555555557, 'direct_attack': 0.4523809523809524, 'undercutter_attack': 0.5172413793103449, 'partial': 0.08}, 'f1': {'support': 0.9244288224956064, 'agreement': 0.4740740740740741, 'direct_attack': 0.4222222222222222, 'undercutter_attack': 0.4225352112676057, 'partial': 0.0851063829787234}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7258771929824561, 'recall': 0.7258771929824561, 'f1': 0.7258771929824561, 'support': None}, 'macro_avg': {'precision': 0.4869190109407501, 'recall': 0.47585039226418535, 'f1': 0.46567334260764637, 'support': None}, 'weighted_avg': {'precision': 0.7253212915040717, 'recall': 0.7258771929824561, 'f1': 0.7113529885417598, 'support': None}}
Loss: tensor(0.8377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, 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(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5622, 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(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2502, 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(5.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2961, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-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(0.2352, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, 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(0.0166, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-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(0.0430, 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(0.0067, 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(0.1268, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5601, 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(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, 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(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6064, 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(9.6608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, 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(0.0003, 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(0.0010, 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(0.0029, 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(9.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5056e-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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.0341, 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(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, 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(0.0220, 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(0.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, 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(0.0193, 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(7.6570e-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(1.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2311e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2106, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0475e-05, 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(0.0002, 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(0.0224, 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(0.8757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4899, 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(0.0047, device='cuda:0', grad_fn=<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(0.0293, 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(0.1163, 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(0.0035, 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(0.0134, 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(0.0026, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, 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(0.0001, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, 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(1.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0100, 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(0.0143, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-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(0.0072, 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(0.0007, 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(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(5.2531e-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(0.0157, 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(0.0008, 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(0.0368, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0031, 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(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0110, device='cuda:0', grad_fn=<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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(4.9746e-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(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, 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(0.0047, 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(7.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8485e-05, 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(0.0001, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0007, 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(0.0030, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1350e-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(3.9694e-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(8.3895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, 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(8.7376e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0011, 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(7.9960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, 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(0.0089, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5688e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, 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(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9949e-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(0.0092, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1937e-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(1.4412, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1432e-05, 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(0.4897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<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(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-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(8.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, 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(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-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(0.0010, 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(4.8050e-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(0.0017, 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(0.0033, 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(0.0016, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3026, 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(0.0004, 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(0.0033, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7397e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1914e-05, 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(0.0089, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0568, 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(0.3336, 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(0.0028, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, 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(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-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(0.0016, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4974e-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(5.2289e-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(0.0229, 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(0.0017, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3654, 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(9.8091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, 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(0.0179, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, 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(0.0022, 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(0.0014, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-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(0.0027, 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(5.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-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(4.9474e-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(0.3079, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-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(0.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6792, 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(3.0066e-05, 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(0.0037, 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(4.9200e-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(0.0039, 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(0.0115, 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(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7525e-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(0.0442, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8036, 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(4.1390e-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(0.0022, 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(0.0035, 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(0.0011, 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(8.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-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(0.0002, 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(0.0027, 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(0.0050, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, 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(0.0002, 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(0.0027, 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(7.7781e-05, device='cuda:0', grad_fn=<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(0.0195, 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(0.0022, 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(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6667e-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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0012, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7551, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3250e-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(0.0003, 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(0.0047, 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(4.4236e-05, device='cuda:0', grad_fn=<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(0.0041, 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(0.0089, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, 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(0.0020, 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(3.3397e-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(0.0080, 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(0.0582, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4104, 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(8.8344e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, 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(0.0020, 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(0.0015, 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(0.0028, 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(0.0012, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0043, 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(3.7636e-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(0.0028, 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(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-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(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, 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(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-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(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0007, 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(0.0016, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2813e-05, 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(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-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(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2550e-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(0.0026, 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(0.0047, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6526e-05, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, 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(0.0022, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1918, 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(0.0014, 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(0.0019, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-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(0.0380, 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(0.0026, 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(0.0011, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-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(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, 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(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, 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(0.0017, 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(3.1247e-05, 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(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-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(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-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(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, 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(0.0027, 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(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2433, 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(0.0020, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4125e-05, 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(9.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, 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(0.0001, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, 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(0.0003, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, 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(0.0149, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8214e-05, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, 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(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, 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(0.0080, 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(0.0120, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7439e-05, 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(9.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4857, 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(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0318, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238, 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(0.0068, 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(0.0033, 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(0.1469, 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(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, 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(0.0067, 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(0.0120, 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(0.8031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1178, 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(0.9328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0061, 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(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4153, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6952, 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(0.0039, 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(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, 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(0.0005, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3520, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, 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(6.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-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(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, 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(3.0157e-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(0.0221, device='cuda:0', grad_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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, 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
				 {'precision': {'support': 0.8729641693811075, 'agreement': 0.6571428571428571, 'direct_attack': 0.5, 'undercutter_attack': 0.38461538461538464, 'partial': 0.5}, 'recall': {'support': 0.9925925925925926, 'agreement': 0.5111111111111111, 'direct_attack': 0.42857142857142855, 'undercutter_attack': 0.5172413793103449, 'partial': 0.08}, 'f1': {'support': 0.92894280762565, 'agreement': 0.575, 'direct_attack': 0.4615384615384615, 'undercutter_attack': 0.4411764705882353, 'partial': 0.13793103448275865}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7653508771929824, 'recall': 0.7653508771929824, 'f1': 0.7653508771929824, 'support': None}, 'macro_avg': {'precision': 0.5829444822278698, 'recall': 0.5059033023170955, 'f1': 0.508917754847021, 'support': None}, 'weighted_avg': {'precision': 0.7445110285736893, 'recall': 0.7653508771929824, 'f1': 0.7416481731418171, 'support': None}}
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, 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(0.0718, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7123e-05, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-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(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, 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(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, 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(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, 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(7.3178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-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(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, 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(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-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(0.0190, 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(0.0135, 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(2.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-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(1.1980, 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(0.3258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, 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(0.0049, 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(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(4.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9979e-05, 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(0.0021, 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(0.0854, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, 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(0.0662, 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(8.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-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(0.0016, 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(0.0017, 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(0.0049, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9207, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, 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(0.0107, device='cuda:0', grad_fn=<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(0.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2215, 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(0.0002, 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(0.0137, 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(0.0009, 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(7.9416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, 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(0.0912, 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(0.0097, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, 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(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-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(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(3.9755e-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(0.0055, 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(0.0063, 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(0.0295, 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(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, 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(0.1311, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6903e-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(0.0017, 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(0.0001, 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(4.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9354e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-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(0.0246, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2598, 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(0.0011, 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(0.0012, 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(7.1816e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3172, 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(0.0274, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, 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(0.2026, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-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(0.0006, 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(0.0001, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0035, 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(0.0035, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, 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(4.0903, device='cuda:0', grad_fn=<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(0.0050, 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(4.3903e-05, 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(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0008, 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(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5552e-05, 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(0.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5001, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, 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(8.0173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, 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(0.0003, 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(6.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, 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(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6360, 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(0.0025, 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(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-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(0.0026, 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(2.8007e-05, 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(0.0019, 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(0.0288, 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(0.0058, 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(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-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(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4277e-05, 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(2.2830e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, 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(9.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, 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(0.0087, 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(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3291e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8543, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, 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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, 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(0.0036, 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(0.0156, 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(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(1.4143, 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(3.0642e-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(1.8167e-05, 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(4.1329e-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(0.0126, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, 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(4.3085e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3306, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9925, 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(0.0004, 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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0066, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, 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(0.0080, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0134, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3221e-05, 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(1.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, 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(0.0257, 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(0.0056, 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(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, 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(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, 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(0.0005, 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(0.0424, 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(0.0358, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, 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(0.0139, 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(0.0104, 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(0.0005, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, 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(0.0003, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, 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(0.0001, 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(0.0053, 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(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, 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(4.9353e-05, 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(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-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(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, 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(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3562e-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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, 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(0.0073, 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(0.0042, 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(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-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(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(3.3669e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, 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(4.1814e-05, 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(0.0007, 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(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8936, 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(0.0008, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, 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(4.9836e-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(0.0028, 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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4651e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, 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(5.2410e-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(0.0054, 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(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, 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(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, 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(8.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7931, 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(0.0398, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2127, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, 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(4.0966e-05, 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(0.0009, 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(0.0103, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9303e-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(0.0545, 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(0.0125, 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(0.0013, 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(3.8332e-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(0.0064, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-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(8.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-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(0.0013, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085e-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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, 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(8.7859e-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(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7798e-05, 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(0.0002, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2338e-05, device='cuda:0', grad_fn=<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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2431, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, 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(0.2717, 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(0.0054, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, 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(0.0060, 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(0.0024, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, 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(0.0017, 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(0.0004, 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(0.0009, 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(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-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(0.4933, 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(0.0033, 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(0.0001, 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(0.0019, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1978e-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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0917e-05, 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(9.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-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(6.5760e-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(3.0127e-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(0.0007, 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(0.1235, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1189e-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(0.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-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(7.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6608e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, 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(0.0004, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, 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(0.0241, 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(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, 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(0.0020, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, 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(0.0041, 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(0.3494, 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(2.7402e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7779, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-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(0.0029, 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(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-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(0.0023, 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(8.9072e-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(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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
				 {'precision': {'support': 0.890728476821192, 'agreement': 0.7142857142857143, 'direct_attack': 0.5172413793103449, 'undercutter_attack': 0.4, 'partial': 0.6}, 'recall': {'support': 0.9962962962962963, 'agreement': 0.5555555555555556, 'direct_attack': 0.35714285714285715, 'undercutter_attack': 0.6896551724137931, 'partial': 0.12}, 'f1': {'support': 0.9405594405594405, 'agreement': 0.6250000000000001, 'direct_attack': 0.4225352112676057, 'undercutter_attack': 0.5063291139240508, 'partial': 0.19999999999999998}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7828947368421053, 'recall': 0.7828947368421053, 'f1': 0.7828947368421053, 'support': None}, 'macro_avg': {'precision': 0.6244511140834502, 'recall': 0.5437299762817005, 'f1': 0.5388847531502194, 'support': None}, 'weighted_avg': {'precision': 0.7743564494703303, 'recall': 0.7828947368421053, 'f1': 0.7623488423861532, 'support': None}}
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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, 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(0.0006, 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(0.0045, 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(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(0.0224, 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(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0001, 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(0.0269, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7182e-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(0.0017, 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(0.0008, 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(0.0005, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6510e-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(7.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0854e-05, 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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-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(0.0009, 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(0.0056, 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(0.0003, 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(0.0045, 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(0.0016, 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(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0274, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, 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(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-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(0.0018, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, 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(0.0910, 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(2.7705e-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(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-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(0.0010, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3420e-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(5.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3076e-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(0.0019, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, 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(0.0018, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7569e-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(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, 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(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5278e-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(7.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-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(0.0106, 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(3.2212, device='cuda:0', grad_fn=<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(0.0031, 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(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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1041e-05, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3437, device='cuda:0', grad_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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3303, 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(4.0815e-05, 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(0.0013, 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(6.9818e-05, 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(1.9378e-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(6.1038e-05, 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(0.0003, 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(6.6186e-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(0.0012, 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(3.1156e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-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(0.0018, 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(0.0011, 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(0.0001, 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(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6782e-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(0.0035, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1684e-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(0.0034, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9020e-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(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, device='cuda:0', grad_fn=<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(0.0044, 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(8.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, 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(0.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-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(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(4.0996e-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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, 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(5.3319e-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(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0021, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, 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(0.0003, 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(0.0026, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-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(0.0510, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, 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(0.0011, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8053e-05, 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(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, 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(0.0014, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, 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(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2733e-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(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, 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(0.0015, 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(0.0044, 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(0.0070, 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(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-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(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-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(0.0020, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0175, 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(5.4196e-05, 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(0.0084, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5820, 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(0.0002, 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(0.0012, 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(0.0018, 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(0.0001, 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(0.0022, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4340, 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(0.0435, 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(1.5328, 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(0.0020, 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(0.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, 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(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0005, 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(0.0002, 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(0.0001, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, 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(0.0007, 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(8.9101e-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(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2651e-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(0.0136, 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(4.9625e-05, 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(0.0015, 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(0.0099, 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(0.0014, 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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(7.7205e-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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3338e-05, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, 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(0.0082, device='cuda:0', grad_fn=<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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-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(8.1474e-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(0.3687, 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(0.0001, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6038, 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(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, 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(0.0058, 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(0.0009, 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(0.0036, 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(0.0011, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7202e-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(5.6498e-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(0.0002, 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(7.4632e-05, 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(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8430, 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(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6247e-05, 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(8.3291e-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(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2495, 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(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-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(0.0016, 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(3.7696e-05, 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(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0010, 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(0.0007, 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(0.0019, 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(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, 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(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-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(0.0100, 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(1.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-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(2.5282e-05, 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(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, 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(5.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, 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(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-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(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-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(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3755, 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(0.0010, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0000e-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(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3397, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-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(3.7423e-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(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, 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(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(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0045, 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(4.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0170, 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(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3561e-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(0.0010, 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(0.0047, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, 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(3.1217e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0044, 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(8.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6002, 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(6.1614e-05, 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(8.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4628, 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(0.1519, 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(0.0585, device='cuda:0', grad_fn=<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(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7355e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0981e-05, 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(0.0014, 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(3.0339e-05, 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(9.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3777, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, 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(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7155e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3290e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2440, 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(9.0128e-05, device='cuda:0', grad_fn=<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(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3060, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-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(0.0008, 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(0.0044, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-05, 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(4.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8806e-05, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2344, 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(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, 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(0.0095, 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(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, 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(0.0022, 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(5.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, 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(9.3371e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5234, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, 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(0.0003, 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(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0034, 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(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, 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(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0623e-05, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7788e-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(9.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.0007, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, 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(0.0003, 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(0.0012, 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(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3034e-05, device='cuda:0', grad_fn=<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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6439, 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(0.0080, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, 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(6.0009e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, 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(3.0763e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, 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(0.0026, device='cuda:0', grad_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(0.0040, 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(0.0105, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0002, 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(6.2067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-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(0.0078, 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(5.7890e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0117, 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(0.0242, 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(0.0032, 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(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(2.8068e-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(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-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(0.0406, 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(8.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(5.2258e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0483e-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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2594e-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(7.2846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, 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(5.9162e-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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0982, 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(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-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
				 {'precision': {'support': 0.8701298701298701, 'agreement': 0.7758620689655172, 'direct_attack': 0.6785714285714286, 'undercutter_attack': 0.37735849056603776, 'partial': 0.4444444444444444}, 'recall': {'support': 0.9925925925925926, 'agreement': 0.5, 'direct_attack': 0.4523809523809524, 'undercutter_attack': 0.6896551724137931, 'partial': 0.16}, 'f1': {'support': 0.9273356401384082, 'agreement': 0.608108108108108, 'direct_attack': 0.5428571428571429, 'undercutter_attack': 0.4878048780487806, 'partial': 0.23529411764705882}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7807017543859649, 'recall': 0.7807017543859649, 'f1': 0.7807017543859649, 'support': None}, 'macro_avg': {'precision': 0.6292732605354596, 'recall': 0.5589257434774677, 'f1': 0.5602799773598997, 'support': None}, 'weighted_avg': {'precision': 0.7792042949111572, 'recall': 0.7807017543859649, 'f1': 0.7630242258151119, 'support': None}}
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1797e-05, 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(0.0006, 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(0.0021, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8475e-05, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-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(6.0917e-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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9009e-05, 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(0.0024, 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(0.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5611e-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(8.6257e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0496, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, 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(0.2489, 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(0.0015, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0022, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0281e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-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(0.0541, 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(0.0023, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, 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(0.0058, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, 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(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(0.0033, 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(0.0023, 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(0.0055, 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(5.9344e-05, 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(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3469e-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(0.0003, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, 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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-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(5.1925e-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(0.0007, 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(4.5991e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, 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(4.7051e-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(0.0024, 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(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, 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(0.0007, 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(0.0001, 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(9.1251e-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(0.0079, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9475, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2106e-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(2.4162e-05, 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(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5864e-05, 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(3.0975e-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(7.9325e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, 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(3.8604e-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(0.0008, 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(0.0039, 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(0.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7011e-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(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6614, 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(1.8440e-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(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4487e-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(0.5827, 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(6.6700e-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(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8931e-05, 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(0.0004, 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(0.0795, 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(0.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3428, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-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(0.0015, 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(0.0016, 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(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-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(6.9999e-05, 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(9.0889e-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(0.0466, 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(0.0542, 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(0.0022, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, 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(4.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6917, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7840, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2485, 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(0.0065, 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(0.0085, 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(0.0144, 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(0.0015, 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(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4470e-05, 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(0.0023, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7032e-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(0.0030, 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(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3902, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1704e-05, 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(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-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(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, 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(5.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, 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(8.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, 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(6.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, 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(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, 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(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2651, device='cuda:0', grad_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(2.1770e-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(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8910, 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(6.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-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(4.5537e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, 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(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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-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(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, 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(0.0109, 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(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4400e-05, 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(0.0153, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-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(0.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(4.6930e-05, 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(0.0004, 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(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, 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(0.0016, 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(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-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(0.1231, 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(8.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-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(0.0015, 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(3.9725e-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(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, 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(0.2458, 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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-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(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, 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(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-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(0.8396, 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(0.0007, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, 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(0.6560, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, 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(3.8967e-05, 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(0.0126, 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(0.0065, 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(0.0091, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, 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(0.0023, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0555e-05, 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(3.9451e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, 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(2.8007e-05, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, 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(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-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(0.0010, 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(7.5509e-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(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-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(0.0068, 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(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.1709, 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(0.0010, 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(2.9188e-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(0.1195, 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(0.0026, 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(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-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(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0164, device='cuda:0', grad_fn=<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(0.0019, 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(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-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(7.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4171, 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(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-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(0.0003, 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(8.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7003e-05, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, 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(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0028, 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(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-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(8.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, 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(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-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(0.0125, 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(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6105, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0054, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-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(0.0028, 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(0.1665, 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(0.1569, 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(0.2262, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7368, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3758, 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(0.0026, 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(0.0011, 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(0.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4072, 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(0.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, 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(4.9897e-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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, 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(0.0001, 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(0.0033, 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(0.0103, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, 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(0.1186, 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(0.2700, 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(0.0012, 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(0.0072, 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(0.0028, 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(0.0017, 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(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, 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(0.0037, 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(0.1205, 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(2.7665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3654, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(8.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0737e-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(0.0004, 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(5.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7799e-05, 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(0.1865, 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(3.6418, 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(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2271, 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(0.0016, 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(0.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9081, 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(0.0035, 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(0.0005, 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(0.0009, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, 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(0.0040, 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(0.0007, 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(0.0064, 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(1.4453, 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(0.0148, 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(0.0083, 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(0.0059, 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(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.2862, 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(0.0080, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0400, 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(5.5693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, 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(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, 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(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8362, device='cuda:0', grad_fn=<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(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, 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
				 {'precision': {'support': 0.8847457627118644, 'agreement': 0.7230769230769231, 'direct_attack': 0.43137254901960786, 'undercutter_attack': 0.3888888888888889, 'partial': 0.18518518518518517}, 'recall': {'support': 0.9666666666666667, 'agreement': 0.5222222222222223, 'direct_attack': 0.5238095238095238, 'undercutter_attack': 0.2413793103448276, 'partial': 0.2}, 'f1': {'support': 0.9238938053097345, 'agreement': 0.6064516129032258, 'direct_attack': 0.4731182795698925, 'undercutter_attack': 0.2978723404255319, 'partial': 0.1923076923076923}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.75, 'recall': 0.75, 'f1': 0.75, 'support': None}, 'macro_avg': {'precision': 0.5226538617764939, 'recall': 0.49081554460864807, 'f1': 0.4987287461032155, 'support': None}, 'weighted_avg': {'precision': 0.7411915207792926, 'recall': 0.75, 'f1': 0.7398002862212433, 'support': None}}
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8394, 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(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, 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(0.0016, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, 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(0.0111, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4521, 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(0.0123, 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(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1776e-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(0.0077, 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(0.9690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8104e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, 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(0.0058, 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(0.5778, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4478, 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(0.0014, 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(0.0325, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0185, 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(0.1036, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.2372, 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(0.0021, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, 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(0.0010, 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(0.0040, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0065, 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(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0813, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3491, 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(0.0112, 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(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0012, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, 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(0.0182, 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(0.0791, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4193, 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(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7504, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, 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(1.3509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, 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(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, 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(0.0020, 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(0.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0075, 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(0.0222, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, 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(0.0526, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, 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(0.0856, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0059, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, 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(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, 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(0.1190, 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(0.0347, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4437, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, 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(1.0105, 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(7.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9061, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, 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(6.3005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, 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(7.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, 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(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, 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(0.0020, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, 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(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-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(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4075e-05, device='cuda:0', grad_fn=<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(0.0467, 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(9.4791e-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(0.0007, 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(0.0025, 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(6.7397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9374, 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(0.0301, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, 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(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, 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(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0435e-05, 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(6.4520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8276e-05, 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(0.0800, 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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, 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(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9102e-05, 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(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, 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(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<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(0.0016, 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(0.2012, 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(0.0109, 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(0.0208, 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(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1441e-05, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, 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(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6097, 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(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, 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(0.0049, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3656, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0055, 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(0.0007, 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(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9455, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, 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(8.4593e-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(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3581e-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(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4153, 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(9.1737e-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(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, 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(0.0001, 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(8.5986e-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(9.1314e-05, 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(0.0122, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, 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(0.0388, 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(0.0002, 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(0.0050, 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(9.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7598e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9394e-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(0.3252, 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(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, 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(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, 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(0.0027, 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(0.0016, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0039, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, 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(0.0653, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4601, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-05, 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(0.0010, 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(8.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, 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(0.0095, 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(0.0017, 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(0.0003, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, 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(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, 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(0.0029, 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(0.0004, 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(0.0061, 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(0.0085, 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(0.0023, 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(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8012e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0513, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0120, 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(0.0350, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-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(0.0029, 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(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-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(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, 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(0.0132, 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(0.0111, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, 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(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, 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(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7729e-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(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, 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(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, 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(0.0020, 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(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2914e-05, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-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(9.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2439e-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(0.0024, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2652e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-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(8.6709e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6376e-05, 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(0.0056, 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(0.6769, 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(0.0026, device='cuda:0', grad_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(0.0168, 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(0.0025, 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(2.7402e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, 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(0.0001, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4321, 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(4.7627e-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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<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(0.0022, 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(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, 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(0.0049, 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(0.1908, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1670, 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(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, 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(0.0001, 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(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, 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(1.2251, 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(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3693e-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(0.0015, 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(0.0001, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, 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(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, 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(0.0049, 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(0.0053, 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(0.0014, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, 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(0.0003, 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(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5501e-05, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-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(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0019, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9848, 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(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, 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(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, 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(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3736e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, 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(7.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2072, 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(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3974e-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(0.0064, 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(0.0044, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, 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(4.4175e-05, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<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(8.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0829e-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(0.5521, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, 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(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, 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(0.0037, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4034, 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(0.0140, 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(0.5391, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6439e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, 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(0.0001, 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(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0715e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2734e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0221e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, 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(0.0002, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8139, 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(0.0327, 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(0.0135, 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(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(5.0432, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0035, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, 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(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-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(0.0046, 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(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-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(9.1101e-05, 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>)
			------------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
				 {'precision': {'support': 0.8758169934640523, 'agreement': 0.7608695652173914, 'direct_attack': 0.4, 'undercutter_attack': 0.3, 'partial': 0.0}, 'recall': {'support': 0.9925925925925926, 'agreement': 0.3888888888888889, 'direct_attack': 0.047619047619047616, 'undercutter_attack': 0.9310344827586207, 'partial': 0.0}, 'f1': {'support': 0.9305555555555556, 'agreement': 0.5147058823529412, 'direct_attack': 0.0851063829787234, 'undercutter_attack': 0.45378151260504207, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7280701754385965, 'recall': 0.7280701754385965, 'f1': 0.7280701754385965, 'support': None}, 'macro_avg': {'precision': 0.46733731173628873, 'recall': 0.47202700237182993, 'f1': 0.39682986669845244, 'support': None}, 'weighted_avg': {'precision': 0.7246685287387266, 'recall': 0.7280701754385965, 'f1': 0.6892711871982836, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2474, 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(0.0919, 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(0.0002, 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(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, 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(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9916e-05, 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(0.0033, 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(0.0001, 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(5.6074e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, 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(0.0010, 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(0.0122, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3710, 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(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, 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(0.0007, 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(0.0042, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-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(0.0205, 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(1.5321e-05, 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(0.0378, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5104, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, 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(0.0358, 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(0.0071, 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(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_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(1.0387, device='cuda:0', grad_fn=<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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9732e-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(9.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8254e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, 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(0.0005, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5384, 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(0.0058, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1893, 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(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.1249, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, 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(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3803, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3272, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, 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(0.0074, 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(0.0015, 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(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3903, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, 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(0.0068, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0234, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, 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(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(1.5887, 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(0.0021, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, 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(0.0054, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, 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(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, 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(0.0203, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, 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(0.0041, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0084, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0021, 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(0.0016, 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(0.0164, 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(0.0033, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0112, 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(0.0116, 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(0.0434, 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(0.9504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, 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(0.0026, 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(0.0035, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, 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(0.0016, 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(0.0087, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, 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(0.3411, 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(0.0029, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, 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(0.0026, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, 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(1.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0016, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, 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(0.0009, 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(0.0104, 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(0.0016, 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(0.0035, 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(0.0022, 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(0.0004, 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(0.0017, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0013, 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(0.0026, 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(0.0026, 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(0.0022, 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(0.0010, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, 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(0.0052, 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(0.0010, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(5.7618e-05, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8240, 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(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, 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(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0008, 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(0.0072, 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(0.0086, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8131, 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(0.0019, 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(0.0015, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, 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(0.1398, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, 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(0.0007, 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(0.0013, 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(0.2856, 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(0.0255, 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(0.0020, 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(0.0094, 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(0.0030, 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(0.0048, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2953, 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(0.0039, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4684, 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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3124, device='cuda:0', grad_fn=<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(0.0006, 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(0.0028, 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(0.0270, 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(0.0019, 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(0.0008, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583, 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(0.0008, 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(0.1659, 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(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, 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(0.0029, 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(0.0015, 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(0.0061, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, 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(0.0041, 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(0.1341, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8122, 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(0.0074, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, 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(0.0084, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, 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(3.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8116, 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(0.0421, 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(0.0033, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, 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(0.0024, 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(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, 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(0.0018, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3774, 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(0.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2683, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1318, 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(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, 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(0.0621, 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(0.0205, 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(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.4550, 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(0.0324, 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(0.0037, 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(0.0075, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5567, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5178, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, 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(1.8263, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8485, 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(0.0017, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3118, 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(0.0009, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4087e-05, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(1.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, 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(0.0001, 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(0.0081, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4002, device='cuda:0', grad_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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, 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(0.0130, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, 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(0.0008, 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(0.0406, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(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(0.0050, device='cuda:0', grad_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(0.0006, 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(0.5112, 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(0.0006, 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(0.5375, 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(0.0018, 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(0.0042, 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(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9493, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5012, 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(0.0001, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(0.0014, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, 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(0.0019, 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(0.0012, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.5095, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, 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(0.0021, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, 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(0.0916, 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(0.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.7354, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8008, 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(0.0049, 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(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6219, 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(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8666, 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(0.0156, 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(0.0009, 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(0.6038, 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(0.0034, 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(0.0616, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, 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(0.0014, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, 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(0.0009, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1728, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, 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(0.0022, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7002, 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(0.4031, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3057, 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(0.1070, 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(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, 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(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5675, 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(0.0001, 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(0.0505, 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(0.2822, 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(0.0211, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, 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(0.0183, 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(0.0011, 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(0.0035, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0200, 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(0.0022, 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(0.0025, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, 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(0.0015, 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(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4083, device='cuda:0', grad_fn=<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(0.0052, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7063, 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(0.0016, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, 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(0.0007, 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(0.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0043, 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(0.0013, 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(0.0008, 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(0.0032, 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(1.1604, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, 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(1.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4880, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5588, 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(0.0087, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, 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(0.2445, 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(0.0007, 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(0.0015, 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>)
			------------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
				 {'precision': {'support': 0.867741935483871, 'agreement': 0.7674418604651163, 'direct_attack': 0.5483870967741935, 'undercutter_attack': 0.4107142857142857, 'partial': 0.25}, 'recall': {'support': 0.9962962962962963, 'agreement': 0.36666666666666664, 'direct_attack': 0.40476190476190477, 'undercutter_attack': 0.7931034482758621, 'partial': 0.16}, 'f1': {'support': 0.9275862068965518, 'agreement': 0.49624060150375937, 'direct_attack': 0.4657534246575342, 'undercutter_attack': 0.5411764705882354, 'partial': 0.19512195121951217}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7587719298245614, 'recall': 0.7587719298245614, 'f1': 0.7587719298245615, 'support': None}, 'macro_avg': {'precision': 0.5688570356874932, 'recall': 0.544165663200146, 'f1': 0.5251757309731186, 'support': None}, 'weighted_avg': {'precision': 0.7555988209928423, 'recall': 0.7587719298245614, 'f1': 0.7351836409223035, 'support': None}}
Loss: tensor(0.7313, 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(0.0262, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, 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(0.0021, 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(0.0045, 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(2.5228, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, 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(0.1171, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0023, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-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(0.0092, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8598e-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(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, 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(0.0020, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<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(5.3137e-05, 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(0.0005, 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(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(6.5822e-05, device='cuda:0', grad_fn=<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(0.0198, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, 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(0.0005, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, 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(0.7384, 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(0.0041, 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(0.0007, 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(6.9698e-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(0.0027, 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(0.0008, 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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, 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(0.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, 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(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3906e-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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0263e-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(0.0005, 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(0.0007, 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(0.0015, 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(0.0003, 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(0.0010, 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(0.0874, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0839, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, 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(0.0400, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3391e-05, 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(0.0486, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, 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(0.0018, 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(0.0008, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3897e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6673, 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(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4474, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, 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(0.0014, 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(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0010, 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(0.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3775e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8233e-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(0.0002, 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(3.3455, 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(8.4892e-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(0.0011, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7489e-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(0.0027, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2494e-05, 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(0.0003, 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(0.0031, 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(0.1554, 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(8.1717e-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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-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(0.0007, 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(6.1826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0005, 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(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-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(0.0005, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4803, device='cuda:0', grad_fn=<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(4.7475e-05, 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(0.0005, 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(0.1002, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-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(0.0127, 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(8.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3728, 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(0.0015, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, 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(3.7484e-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(0.0016, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7377e-05, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6054, 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(0.0009, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8525e-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(0.0001, 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(0.0036, 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(0.0022, 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(0.0112, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, 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(9.0615e-05, 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(6.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7872e-05, 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(0.0006, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, 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(0.0015, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3497, 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(0.0364, 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(0.0017, 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(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, 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(0.0066, 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(0.0002, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1401e-05, 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(0.0020, 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(6.5324, 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(6.9334e-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(0.0006, 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(0.0003, 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(0.0010, 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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1333e-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(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(5.9979e-05, 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(0.0048, 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(7.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-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(0.0321, 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(0.1545, 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(0.0017, 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(0.0043, device='cuda:0', grad_fn=<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(0.0007, 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(0.4610, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, 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(0.0004, 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(0.0034, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2340e-05, 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(3.1550e-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(0.0066, 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(0.0035, 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(4.1450e-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(0.0013, 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(0.0444, 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(0.0009, 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(0.0011, 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(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, 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(0.0001, 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(8.6530e-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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(8.1049e-05, 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(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, 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(0.0071, 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(0.0061, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0004, 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(0.0016, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3074, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1290e-05, device='cuda:0', grad_fn=<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(0.1732, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5292, 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(0.0001, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, 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(0.0018, 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(0.0225, 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(7.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, 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(0.0002, 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(0.0007, 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(0.0024, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, 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(0.0307, 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(0.0054, 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(7.6234e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0006, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7363e-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(0.0107, 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(5.4802e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8840e-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(0.0785, 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(0.0014, 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(0.0013, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6305e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.1400, 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(3.2458e-05, 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(4.8868e-05, device='cuda:0', grad_fn=<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(2.6705e-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(0.0013, 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(0.0005, 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(0.0007, 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(0.0108, 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(0.0015, 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(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, 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(3.3881e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2863, 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(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, 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(8.7590e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5567e-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(8.3501e-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(0.1117, 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(0.0023, 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(0.0430, 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(0.0009, 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(0.8618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, 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(4.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, 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(0.0002, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-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(7.0665e-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(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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8527, device='cuda:0', grad_fn=<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(0.0055, 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(4.1965e-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(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, 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(0.0013, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6971e-05, 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(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9899e-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(2.3950e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3810e-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(0.0015, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9766e-05, 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(0.0018, 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(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(0.0008, 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(6.8396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8818e-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(0.0019, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3035e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2332, 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(7.7084e-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(5.3621e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-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(0.1462, 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(0.0016, 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(0.0003, 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(3.9641e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1254, 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(0.0004, 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(0.0003, 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(4.5719e-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(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3289, 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(0.0022, 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(0.0198, 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(6.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7472, device='cuda:0', grad_fn=<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(0.0018, 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(0.0005, 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(6.1130e-05, device='cuda:0', grad_fn=<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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8937e-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(9.0461e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1915, device='cuda:0', grad_fn=<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(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6532, 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(0.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5134e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, 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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, 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(0.0424, 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(0.0167, 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(0.0723, 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(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, 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(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(0.0022, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4803, 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(0.0060, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, 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(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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
				 {'precision': {'support': 0.8952702702702703, 'agreement': 0.7246376811594203, 'direct_attack': 0.4666666666666667, 'undercutter_attack': 0.34615384615384615, 'partial': 0.4444444444444444}, 'recall': {'support': 0.9814814814814815, 'agreement': 0.5555555555555556, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.6206896551724138, 'partial': 0.16}, 'f1': {'support': 0.9363957597173145, 'agreement': 0.628930817610063, 'direct_attack': 0.3888888888888889, 'undercutter_attack': 0.4444444444444444, 'partial': 0.23529411764705882}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7697368421052632, 'recall': 0.7697368421052632, 'f1': 0.7697368421052632, 'support': None}, 'macro_avg': {'precision': 0.5754345817389297, 'recall': 0.5302120051085569, 'f1': 0.5267908056615539, 'support': None}, 'weighted_avg': {'precision': 0.7624779318572225, 'recall': 0.7697368421052632, 'f1': 0.7555596576139896, 'support': None}}
Loss: tensor(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, 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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, 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(0.0113, 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(0.0002, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-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(0.4887, 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(0.0013, 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(0.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3033, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, 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(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-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(4.8928e-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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0023, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, 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(0.0011, 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(7.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0493e-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(0.0002, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, 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(0.0239, device='cuda:0', grad_fn=<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(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, 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(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-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(0.0026, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, 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(0.1308, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8731e-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(5.4347e-05, device='cuda:0', grad_fn=<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(8.5830e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4651, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2407e-05, 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(0.0007, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5332e-05, 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(0.0010, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, 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(1.0395, 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(4.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9657e-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(5.2955e-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(0.0007, 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(0.0937, 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(0.0247, device='cuda:0', grad_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(5.2198e-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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2577, 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(0.0002, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3374, device='cuda:0', grad_fn=<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(0.0508, 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(9.0371e-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(9.4155e-05, device='cuda:0', grad_fn=<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(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5969, 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(0.0116, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, 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(0.9964, 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(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6056e-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(0.0305, 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(0.0002, 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(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8076, 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(5.2533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0005, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(8.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, 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(0.0026, 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(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-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(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, 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(0.0003, 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(4.7050e-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(0.0026, 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(6.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7445e-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(0.0100, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(2.5101e-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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3367e-05, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4360e-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(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.6719, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-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(3.7605e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9665e-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(0.0024, 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(0.0073, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, 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(0.0025, 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(0.0007, 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(9.9940e-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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-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(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5534e-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(5.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2469e-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(0.0013, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0008, 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(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, 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(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1337e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, 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(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1284e-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(3.0490e-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(0.0110, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-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(0.0004, 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(9.2071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-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(3.7363e-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(0.0003, 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(0.0001, 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(0.0069, 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(0.0001, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-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(0.0020, 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(0.0003, 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(0.0060, 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(0.0015, 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(5.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(7.5874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8692e-05, 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(0.0002, 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(0.0009, 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(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8994e-05, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, 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(0.0012, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3995, 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(0.0393, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-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(2.5161e-05, 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(6.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, 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(6.9513e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9656e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-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(8.1606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(8.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5669, 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(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-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(0.0007, 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(0.0178, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0001, 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(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8979e-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(7.2420e-05, device='cuda:0', grad_fn=<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(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9907e-05, 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(3.8725e-05, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-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(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-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(0.0001, 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(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, 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(0.0005, 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(0.0119, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, 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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-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(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7266e-05, 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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7265e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0002, 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(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4563e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2754e-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(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0002, 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(0.0012, 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(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, 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(3.2398e-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(0.5028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-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(4.7565e-05, 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(0.0051, 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(0.8453, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5639e-05, 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(0.0009, 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(0.0021, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5347e-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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, 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(2.0641, 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(5.9200e-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(0.0073, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0002, 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(0.0002, 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(9.3584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<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(0.0017, 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(0.0013, 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(0.0010, 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(9.6635e-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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(1.4110e-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(5.3046e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2640e-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(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4607e-05, 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(3.7118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, 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(0.0001, 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(0.0044, 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(0.0009, 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(0.0022, 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(0.0020, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-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(1.5868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4315e-05, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8167e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-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(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<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(8.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7607e-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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7105e-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(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(0.0028, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, 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(5.6317e-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(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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-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(1.9499e-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(2.6251e-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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6558e-05, 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(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, 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(8.7526e-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(6.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, 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(0.0004, 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(9.4367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, 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(6.2944e-05, device='cuda:0', grad_fn=<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(2.3617e-05, 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(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-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(0.0360, device='cuda:0', grad_fn=<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(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-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(4.0238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, 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(0.0001, 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(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-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(2.0317e-05, 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(0.0039, 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(3.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, 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(0.0018, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4953e-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(9.4704e-05, 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(3.5425e-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(0.0008, 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(0.0017, 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(8.9431e-05, 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(6.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, 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(5.1290e-05, 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(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(8.2866e-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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2030e-05, 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(0.0003, 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(0.0643, 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(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7611, 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(0.0007, 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(4.3720e-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(0.0021, 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(0.0006, 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(0.0056, 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(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0013, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0016, 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(0.0019, 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(4.6203e-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(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4371e-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
				 {'precision': {'support': 0.8874172185430463, 'agreement': 0.7543859649122807, 'direct_attack': 0.42592592592592593, 'undercutter_attack': 0.40476190476190477, 'partial': 0.0}, 'recall': {'support': 0.9925925925925926, 'agreement': 0.4777777777777778, 'direct_attack': 0.5476190476190477, 'undercutter_attack': 0.5862068965517241, 'partial': 0.0}, 'f1': {'support': 0.9370629370629371, 'agreement': 0.5850340136054423, 'direct_attack': 0.4791666666666667, 'undercutter_attack': 0.4788732394366197, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7697368421052632, 'recall': 0.7697368421052632, 'f1': 0.7697368421052632, 'support': None}, 'macro_avg': {'precision': 0.4944982028286315, 'recall': 0.5208392629082285, 'f1': 0.4960273713543331, 'support': None}, 'weighted_avg': {'precision': 0.7393078288941051, 'recall': 0.7697368421052632, 'f1': 0.7448955661735631, 'support': None}}
Loss: tensor(0.0013, 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(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(5.9433e-05, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, 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>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5558e-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(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-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(0.0510, 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(0.0001, 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(9.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-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(8.3800e-05, 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(1.8924e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-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(0.0009, 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(0.0024, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7747e-05, 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(0.0009, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, 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(6.0100e-05, 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(0.0001, 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(0.0020, 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(0.2872, 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(2.5949e-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(0.0002, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0038, 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(3.1974e-05, device='cuda:0', grad_fn=<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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-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(0.4115, 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(3.7757e-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(0.0016, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2025e-05, 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(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, 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(0.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1460, 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(6.0432e-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(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0007, 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(0.0002, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(6.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.8302e-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(5.0372, 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(6.0371e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-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(0.0187, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, 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(0.0014, 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(7.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-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(6.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(0.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, 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(8.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2412, 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(3.4366e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9070e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, 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(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, 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(0.1165, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, 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(3.5910e-05, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8956, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6211e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, 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(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-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(3.0823e-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(0.0010, 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(4.8746e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7909e-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(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2317e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3947, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, 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(5.8072e-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(0.0041, 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(3.1520e-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(0.0711, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3596, 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(1.8016e-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(0.0003, 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(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, 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(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7343e-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(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-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(0.0004, 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(2.3829e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8959e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2877, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-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(0.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2682e-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(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<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(0.1370, 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(0.0304, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9842, 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(0.0002, 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(4.7630, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.9802e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7366e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4854, 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(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4255e-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(0.0043, device='cuda:0', grad_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(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, 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(0.0005, 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(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0012, 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(4.7540, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5229e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2614, 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(0.0175, 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(0.0005, 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(7.3573e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, 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(0.0052, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, 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(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(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(0.0004, 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(0.0014, 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(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, 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(0.0170, 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(0.0004, 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(0.0005, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, 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(0.0006, 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(0.0057, 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(0.0022, 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(0.0036, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5024e-05, 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(0.0021, 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(0.0133, 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(0.0053, 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(0.0035, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0012, 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(0.0037, 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(3.3275e-05, 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(0.0001, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0062, device='cuda:0', grad_fn=<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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, 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(4.8505e-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(0.0040, 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(9.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5821e-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(0.0011, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-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(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0471, 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(1.9742e-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(0.0002, 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(2.6614e-05, 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(0.0013, 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(6.5789e-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(6.8545e-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(1.4953, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, 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(2.0105e-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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, 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(4.5567e-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(0.0009, 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(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4357e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4094e-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(9.0478e-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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(1.9015e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, 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(1.6774e-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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0138e-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(0.0031, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7646e-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(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-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(5.1258e-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(3.4729e-05, 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(0.0049, 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(0.0009, 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(0.0002, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0059, 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(0.0001, 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(2.4980e-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(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<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(0.0018, 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(5.7768e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1845e-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(0.0009, 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(1.4292e-05, device='cuda:0', grad_fn=<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(0.0012, 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(7.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, 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(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4879e-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(0.1643, 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(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0438e-05, 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(0.0025, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-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(0.0330, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(6.5669e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8366e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, 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(0.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, 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(0.0019, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-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(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<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(3.0187e-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(0.0105, 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(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0142e-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(0.2127, 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(0.9903, 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(0.0017, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8745e-05, 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(2.5282e-05, 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(1.8319e-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(0.0003, 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(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(4.7536e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2893, 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(1.7017e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8785e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-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(0.0086, 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(0.0155, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(9.0885e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7879e-05, 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(0.0007, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0289, 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(0.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-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(3.0975e-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(0.0018, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9807e-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(0.0020, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3504e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(7.3180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4804, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8609e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1984e-05, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, 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(0.0002, 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(0.1474, 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(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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6769e-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(0.0034, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-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(1.0658e-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(0.0005, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, 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(0.0450, 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(8.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9050, 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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5204e-05, 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(0.0963, 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(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, 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(0.0587, 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(0.0090, 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(3.3003e-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(0.0404, 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(0.0010, 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(6.1735e-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(0.0012, 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(0.0025, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, 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(0.0043, 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(0.0024, 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(0.0327, 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(0.0038, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9960e-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(0.2757, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8767e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-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(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, 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(0.0004, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, 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(0.0011, 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
				 {'precision': {'support': 0.893687707641196, 'agreement': 0.7666666666666667, 'direct_attack': 0.42105263157894735, 'undercutter_attack': 0.34285714285714286, 'partial': 0.0}, 'recall': {'support': 0.9962962962962963, 'agreement': 0.5111111111111111, 'direct_attack': 0.5714285714285714, 'undercutter_attack': 0.41379310344827586, 'partial': 0.0}, 'f1': {'support': 0.9422066549912435, 'agreement': 0.6133333333333334, 'direct_attack': 0.48484848484848486, 'undercutter_attack': 0.375, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7697368421052632, 'recall': 0.7697368421052632, 'f1': 0.7697368421052632, 'support': None}, 'macro_avg': {'precision': 0.4848528297487906, 'recall': 0.498525816456851, 'f1': 0.48307769463461236, 'support': None}, 'weighted_avg': {'precision': 0.7410586595006488, 'recall': 0.7697368421052632, 'f1': 0.7474439324808598, 'support': None}}
Loss: tensor(0.0002, 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(0.0032, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, 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(9.2040e-05, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9229e-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(0.0037, 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(0.0065, device='cuda:0', grad_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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, 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(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, 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(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(2.5858e-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(3.3699e-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(0.0004, 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(0.0054, 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(0.0047, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0957e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-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(0.0002, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0007, 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(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1622e-05, 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(3.2912e-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(2.5646e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9209e-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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7884e-05, 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(0.0058, 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(0.0011, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2695, 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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-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(0.0006, 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(0.1344, device='cuda:0', grad_fn=<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(0.0011, 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(0.0810, 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(0.0041, 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(0.0019, 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(0.0004, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, 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(0.0046, 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(0.0047, 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(0.0213, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, 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(0.0088, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, 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(0.0049, 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(0.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4713, 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(0.0120, 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(0.0094, 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(0.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3853, 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(0.0012, 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(0.4096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, 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(0.0026, 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(7.4236e-05, 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(0.0001, 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(0.0011, 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(2.6887e-05, 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(0.0001, 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(9.5004e-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(0.0005, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, 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(4.9201e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0057, 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(0.0006, 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(0.0004, 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(0.0001, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, 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(0.0128, device='cuda:0', grad_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(0.1201, 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(0.0069, device='cuda:0', grad_fn=<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(0.0021, 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(0.0007, 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(0.0002, 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(2.3557e-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(3.1685, 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(0.0106, 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(0.0014, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3541e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5836, 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(0.0013, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6640e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, 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(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(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, 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(0.0009, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, 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(6.7447e-05, device='cuda:0', grad_fn=<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(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2995e-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(0.0002, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(7.2511e-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(0.0003, 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(3.2579e-05, 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(0.0117, device='cuda:0', grad_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(0.0207, device='cuda:0', grad_fn=<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(0.0012, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, 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(1.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, 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(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, 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(9.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5115e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, 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(5.0079e-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(6.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4853e-05, 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(0.0001, 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(0.0048, 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(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, 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(0.0004, 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(0.0002, 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(0.0136, 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(0.0019, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-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(0.0016, 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(0.0195, 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(2.4192e-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(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<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(0.0021, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, 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(0.0011, 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(0.0008, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-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(0.0005, 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(3.3366e-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(0.1985, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-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(0.0014, 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(0.0006, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0005, 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(0.0039, 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(0.0001, 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(0.0003, 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(0.0011, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0656, 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(0.0030, 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(0.0002, 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(7.9873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5037e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, 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(0.0078, 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(0.0047, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<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(5.4742e-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(0.0010, 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(6.6419, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, 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(0.0013, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3396, device='cuda:0', grad_fn=<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(0.0015, 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(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-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(0.0004, 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(0.0014, 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(5.8859e-05, 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(0.0004, 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(0.0045, 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(0.0010, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(4.4175e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4470, 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(3.7848e-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(0.6188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4008e-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(0.0057, 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(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<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(0.0026, 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(0.0090, 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(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, 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(0.0003, 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(0.0046, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4874e-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(5.6588e-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(0.0005, 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(0.0001, 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(2.3950e-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(0.0012, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-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(0.0536, 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(0.0003, 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(0.0066, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2018e-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(0.0045, 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(0.0016, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, 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(0.0008, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0043, 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(0.0007, 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(0.0057, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2406, 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(0.0029, 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(0.0062, 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(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0020, 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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, 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(7.2301e-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(0.0102, 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(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1645e-05, 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(0.0054, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4078e-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(8.8104e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-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(0.0029, 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(0.6492, device='cuda:0', grad_fn=<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(0.0003, 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(0.0019, 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(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8095e-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(0.0045, 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(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-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(0.0005, 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(0.1068, 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(5.9071e-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(0.0015, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0345e-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(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(0.0001, 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(0.3427, 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(0.0021, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, 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(0.0006, 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(0.0004, 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(0.0010, 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(8.2319e-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(0.0005, 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(0.0146, 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(0.0003, 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(0.0039, 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(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, 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(0.0003, 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(0.0003, 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(3.6122e-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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0001, 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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, 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(0.0025, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8393e-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(0.0053, 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(5.6164e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, 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(0.0028, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3496, device='cuda:0', grad_fn=<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(0.0011, 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(0.9675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(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(7.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, 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(0.0953, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, 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(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.1096, 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(0.0061, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, 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(0.3269, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0032, 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(7.0969e-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(0.0007, 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(0.0003, 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(0.0002, 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(0.0012, 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(0.1734, 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(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, 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(0.0827, 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(0.0004, 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(0.7211, 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(2.2877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3508, 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(2.8583e-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(9.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-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(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(6.5942e-05, 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(0.0025, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, 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(0.0016, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0608, device='cuda:0', grad_fn=<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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7461, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0013, 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(9.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-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(0.0007, 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(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1507e-05, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(1.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6344e-05, 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(0.4394, 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(0.0017, 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(0.2287, 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(0.0014, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, 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(8.4531e-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(0.0500, 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(0.0027, 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
				 {'precision': {'support': 0.8841059602649006, 'agreement': 0.8269230769230769, 'direct_attack': 0.30952380952380953, 'undercutter_attack': 0.3148148148148148, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9888888888888889, 'agreement': 0.4777777777777778, 'direct_attack': 0.30952380952380953, 'undercutter_attack': 0.5862068965517241, 'partial': 0.08}, 'f1': {'support': 0.9335664335664335, 'agreement': 0.6056338028169014, 'direct_attack': 0.30952380952380953, 'undercutter_attack': 0.4096385542168675, 'partial': 0.12903225806451613}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.75, 'recall': 0.75, 'f1': 0.75, 'support': None}, 'macro_avg': {'precision': 0.5337401989719871, 'recall': 0.4884794745484401, 'f1': 0.4774789716377056, 'support': None}, 'weighted_avg': {'precision': 0.7534970376262347, 'recall': 0.75, 'f1': 0.7339370698253516, 'support': None}}
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(2.4919e-05, 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(0.0036, 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(0.0007, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7870, 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(0.0003, 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(0.0024, device='cuda:0', grad_fn=<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(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8612, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-05, 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(0.0727, 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(0.0064, 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(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0024, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4754e-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(0.0026, 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(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0744e-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(0.0099, device='cuda:0', grad_fn=<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(0.0015, 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(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, 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(0.0009, 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(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, 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(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0212, 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(0.2170, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5155e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0470e-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(0.0022, 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(0.0049, 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(0.2152, 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(0.4654, 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(0.0017, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2463e-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(0.0009, 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(0.0004, 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(8.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, 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(0.0883, 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(0.2886, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, 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(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.2669, 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(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, 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(0.0020, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1987e-05, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, 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(0.0112, 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(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(1.3958e-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(0.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-05, 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(0.0436, 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(8.1106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, 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(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0010, 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(0.0003, 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(0.0001, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, 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(0.0079, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<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(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, 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(0.0012, 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(0.0001, 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(4.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, 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(1.9166e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, 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(0.0015, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, 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(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, 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(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3731, device='cuda:0', grad_fn=<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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0189, device='cuda:0', grad_fn=<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(0.0008, 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(0.0011, 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(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3163, 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(1.2126, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.4622, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, 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(1.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, 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(0.0307, 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(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.0438, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0004, 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(8.2351e-05, 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(0.0008, 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(0.0056, 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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3416, 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(0.0203, 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(0.0215, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6279, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-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(1.3242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8356e-05, device='cuda:0', grad_fn=<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(0.0017, 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(3.2670e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3147, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0947e-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(0.0098, 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(0.0367, 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(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-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(0.0296, 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(0.0463, 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(0.0042, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, 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(0.0020, 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(0.1263, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0015, 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(0.0111, 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(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-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(0.6898, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-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(0.0025, 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(0.0001, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, 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(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(2.2891e-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(0.0029, 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(0.0092, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-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(0.0122, 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(1.6828, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, 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(4.7263e-05, 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(3.9573e-05, 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(0.0226, 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(4.5113e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8558, 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(2.7190e-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(0.0010, 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(0.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, 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(0.0007, 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(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, 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(0.5983, device='cuda:0', grad_fn=<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(0.0020, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0021, 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(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, 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(0.0608, 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(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-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(5.6558e-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(0.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0010, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-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(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6165e-05, 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(6.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8183, 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(7.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-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(0.0485, 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(9.7740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1198e-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(0.0068, 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(0.0004, 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(0.0117, 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(0.0371, 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(0.0024, 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(0.0093, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0011, 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(8.2651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(6.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4059e-05, 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(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1776e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-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(0.0003, 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(0.0013, 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(0.0018, 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(0.0015, device='cuda:0', grad_fn=<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(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, 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(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, 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(0.0002, 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(0.0116, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, 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(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, 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(0.0083, 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(0.0034, 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(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, 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(1.0732, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, 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(0.0017, 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(0.0045, 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(1.4443e-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(0.0048, 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(0.0012, 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(3.6212e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, 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(0.9373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, 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(5.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, 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(0.0005, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-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(0.0029, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8283e-05, 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(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0017e-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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0510, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, 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(0.0323, 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(0.0145, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4328, 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(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1561e-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(0.0055, 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(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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0023, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, 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(0.0050, 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(0.0013, 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(0.0023, 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(0.0008, 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(0.0022, 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(0.0071, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, 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(0.0038, 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(0.0046, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, 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(0.0009, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7837, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2292e-05, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4512, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(1.9411, 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(0.0002, 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(0.2339, 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(0.0028, 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(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, 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(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5732e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5929, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(8.6551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, 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(0.0016, 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(0.0002, 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(0.0009, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9467, device='cuda:0', grad_fn=<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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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
				 {'precision': {'support': 0.88, 'agreement': 0.7758620689655172, 'direct_attack': 0.45454545454545453, 'undercutter_attack': 0.3582089552238806, 'partial': 0.0}, 'recall': {'support': 0.9777777777777777, 'agreement': 0.5, 'direct_attack': 0.23809523809523808, 'undercutter_attack': 0.8275862068965517, 'partial': 0.0}, 'f1': {'support': 0.9263157894736842, 'agreement': 0.608108108108108, 'direct_attack': 0.31249999999999994, 'undercutter_attack': 0.5, 'partial': 0.0}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7521929824561403, 'recall': 0.7521929824561403, 'f1': 0.7521929824561403, 'support': None}, 'macro_avg': {'precision': 0.4937232957469705, 'recall': 0.5086918445539135, 'f1': 0.4693847795163585, 'support': None}, 'weighted_avg': {'precision': 0.738830164472145, 'recall': 0.7521929824561403, 'f1': 0.7290789317711063, 'support': None}}
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2263, 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(0.0004, 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(0.1944, 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(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7038e-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(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5428e-05, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4233, 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(0.3564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3922, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-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(0.0013, 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(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<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(0.4049, 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(8.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(3.9002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, 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(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, 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(0.0002, 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(9.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3057, 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(0.0024, 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(0.0002, 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(8.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6475, 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(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0148, 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(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-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(2.9612e-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(0.0154, 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(0.0049, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, 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(8.5647e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1735e-05, 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(6.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2926, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8302e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, 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(8.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(3.4002e-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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662, 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(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-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(0.1654, device='cuda:0', grad_fn=<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(0.4353, device='cuda:0', grad_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(0.0028, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, 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(0.0008, 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(0.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, 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(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7098, device='cuda:0', grad_fn=<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(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, 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(0.0244, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, 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(0.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4969, 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(0.0330, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0484, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5431, 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(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>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5006, device='cuda:0', grad_fn=<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(8.3834e-05, device='cuda:0', grad_fn=<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(0.0015, 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(3.6182e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-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(1.2840, 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(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1812, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<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(3.2731e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, 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(1.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, 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(0.0014, 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(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(9.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9221, 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(0.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, 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(0.0006, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, 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(0.0003, 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(0.0035, 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(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8912, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0027, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8263e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, 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(0.0001, 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(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3928, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, 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(7.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-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(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(0.0340, 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(9.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, 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(0.0002, 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(0.4011, 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(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, 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(0.0005, 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(0.6176, 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(0.0011, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7666, 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(0.0120, 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(0.0595, 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(9.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, 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(2.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916, 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(0.0171, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3334, 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(0.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, 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(0.0015, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4136, 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(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(0.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5724, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0546e-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(0.0013, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8752, 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(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, 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(0.4994, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, 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(1.8314, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1864, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4591, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, 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(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, 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(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0600, 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(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, 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(0.0094, 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(0.0062, 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(0.0035, 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(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6527, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0048, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, 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(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7779, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0030, 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(3.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, 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(5.8039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, 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(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7626, 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(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3879, 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(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2317, 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(0.0076, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2346, 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(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, 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(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, 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(0.0644, 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(0.1517, 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(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3483, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, 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(0.5776, 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(0.0020, 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(0.0006, 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(0.0026, 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(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, 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(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, 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(0.4422, 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(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4414, device='cuda:0', grad_fn=<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(0.0019, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2370, 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(0.0034, 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(0.0067, device='cuda:0', grad_fn=<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(0.0031, 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(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0005, 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(0.0003, 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(0.0035, 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(0.0315, 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(0.0001, 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(0.0004, 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(0.0628, 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(0.0443, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6882e-05, device='cuda:0', grad_fn=<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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3588, 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(0.0003, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, 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(0.0037, 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(0.0082, 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(0.0830, 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(0.0026, 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(0.0079, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, 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(0.0004, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9207, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0017, 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(0.0052, 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(0.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, 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(0.0264, 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(9.6884e-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(0.0431, 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(0.0012, 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(0.0012, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, 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(0.0019, 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(0.0863, 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(0.0006, 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(0.1749, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, 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(0.0044, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<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(0.0007, 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(0.0010, 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(0.0024, 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(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(0.0005, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, 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(0.0010, 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(0.0002, 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(0.0215, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8228, 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(0.0060, 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(0.0010, 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(0.0013, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, 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(0.0003, 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(0.0072, 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(0.0341, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, 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(0.1049, device='cuda:0', grad_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(0.0003, 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(0.0006, 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(0.0020, 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(0.0010, 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(0.0004, 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(0.0003, 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>)
			------------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
				 {'precision': {'support': 0.826625386996904, 'agreement': 0.7631578947368421, 'direct_attack': 0.4222222222222222, 'undercutter_attack': 0.32, 'partial': 0.08}, 'recall': {'support': 0.9888888888888889, 'agreement': 0.32222222222222224, 'direct_attack': 0.4523809523809524, 'undercutter_attack': 0.27586206896551724, 'partial': 0.08}, 'f1': {'support': 0.9005059021922428, 'agreement': 0.453125, 'direct_attack': 0.4367816091954023, 'undercutter_attack': 0.29629629629629634, 'partial': 0.08}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7127192982456141, 'recall': 0.7127192982456141, 'f1': 0.7127192982456141, 'support': None}, 'macro_avg': {'precision': 0.48240110079119364, 'recall': 0.4238708264915162, 'f1': 0.4333417615367883, 'support': None}, 'weighted_avg': {'precision': 0.7036982419930115, 'recall': 0.7127192982456141, 'f1': 0.6860861047603182, 'support': None}}
Loss: tensor(0.0017, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.0063, 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(0.0042, 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(0.0294, 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(0.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0025, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, 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(0.0147, 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(0.0037, 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(6.1572, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7712, 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(0.0016, 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(0.0005, 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(0.0012, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, 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(0.0006, 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(0.0011, 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(0.0003, 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(0.1022, 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(0.0028, 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(0.0036, 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(0.0012, 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(0.0027, 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(0.0009, 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(0.0014, 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(0.4586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, 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(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(0.0022, 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(0.0008, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, 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(0.0012, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, 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(0.0426, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, 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(0.0085, 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(0.0020, 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(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0093, 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(0.0175, 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(0.0014, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476, 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(0.0028, 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(0.1437, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0072, 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(0.0010, 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(0.0015, 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(0.0020, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0055, 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(0.0031, device='cuda:0', grad_fn=<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(0.0026, 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(0.0010, 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(0.0026, 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(0.0004, 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(0.0046, 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(0.0006, 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(0.0036, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(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(0.0014, 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(0.0029, 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(0.0357, 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(0.0079, 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(0.0053, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, 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(0.0028, 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(0.0006, 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(0.0008, 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(0.0001, 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(0.0011, 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(0.0020, 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(0.0008, 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(0.0013, 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(0.0011, 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(0.0269, 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(0.0005, 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(0.0011, 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(0.0008, 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(0.0125, 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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3502, 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(0.0003, 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(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, 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(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(0.0015, 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(0.8502, 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(0.0213, 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(0.0013, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<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(0.0041, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0010, 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(0.0012, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0008, 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(0.0011, 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(0.0012, 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(0.0030, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, 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(0.0018, 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(0.0008, 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(0.0041, 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(0.0132, 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(0.0325, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1521, 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(0.0010, 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(0.0056, 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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0113, 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(0.0015, 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(0.0286, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.0574, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, 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(0.0096, 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(0.0016, 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(0.0007, 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(0.0035, 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(0.0032, 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(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(0.0006, 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(0.0020, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3568, 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(0.0022, 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(0.0021, 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(0.0014, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3330e-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(0.0175, 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(0.0005, 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(0.0345, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0055, 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(0.0036, 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(0.0012, 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(0.0007, 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(0.0001, 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(0.0002, 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(0.0003, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<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(0.0055, 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(0.0020, 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(0.0253, 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(0.0019, 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(0.0003, 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(0.0018, 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(0.0408, 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(0.0003, 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(0.0004, 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(8.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0056, 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(10.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0604, 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(0.0274, 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(0.0023, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0039, 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(0.0002, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, 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(0.0002, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, 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(0.0008, 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(0.0036, 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(0.0028, 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(0.0216, device='cuda:0', grad_fn=<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(0.0023, 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(0.0009, 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(0.0065, 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(0.0299, 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(0.0665, 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(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0020, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6133, 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(0.0001, 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(0.0060, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, 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(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, 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(0.1297, device='cuda:0', grad_fn=<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(2.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, 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(0.0007, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3306, 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(0.3979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0253, 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(0.0006, 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(0.0816, 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(0.0005, 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(0.0002, 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(0.0004, 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(0.0003, 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(0.0064, 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(0.3857, 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(1.2904, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5088, 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(0.0174, 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(0.0023, 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(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, 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(0.2543, 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(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.1492, 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(0.0043, 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(9.8008e-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(0.0008, 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(0.0001, 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(0.0120, 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(0.0008, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, 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(0.0011, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, 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(0.0081, 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(0.0624, 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(0.0163, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2212, 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(8.9770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4356, 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(0.0269, 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(0.0011, 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(0.0034, 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(0.0014, 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(0.0362, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3951, 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(0.0378, device='cuda:0', grad_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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0118, 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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5140, 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(0.0063, 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(0.0039, 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(9.4701e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4659, 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(0.0020, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(9.9154e-05, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, 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(1.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<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(0.0067, 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(0.0036, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<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(0.0006, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.0237, 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(0.0006, 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(0.0002, 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(0.0086, device='cuda:0', grad_fn=<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(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1382e-05, 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(0.0085, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-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(8.0747e-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(0.0001, 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(7.3997e-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(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, 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(6.6852e-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(7.1393e-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(0.0171, 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(1.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-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(0.0022, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0001, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-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(0.0029, 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(1.6672, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0019, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9386e-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(0.0001, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, 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(0.0023, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, 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(0.0002, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1271e-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(0.0011, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3187, 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(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6292, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, 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(7.5868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(8.1957e-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(0.0015, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6950e-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(0.0053, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8871, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, 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(0.1670, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6336e-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(0.0010, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, 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(0.0002, 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(0.0006, 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(9.5760e-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(0.0173, device='cuda:0', grad_fn=<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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(1.8770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3431e-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(5.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(8.1747e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1111e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.3492, 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(0.0001, 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(0.0003, 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(0.0001, 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(0.0001, 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(0.0026, 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(9.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, 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(2.4162e-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(0.0005, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-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(0.0015, 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(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(0.0510, 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(9.7453e-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(0.0069, 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(0.0128, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-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(0.8800, 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(9.1495e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-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
				 {'precision': {'support': 0.8653846153846154, 'agreement': 0.803921568627451, 'direct_attack': 0.5454545454545454, 'undercutter_attack': 0.36363636363636365, 'partial': 0.3125}, 'recall': {'support': 1.0, 'agreement': 0.45555555555555555, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.8275862068965517, 'partial': 0.2}, 'f1': {'support': 0.9278350515463918, 'agreement': 0.5815602836879433, 'direct_attack': 0.22641509433962262, 'undercutter_attack': 0.5052631578947369, 'partial': 0.24390243902439027}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.7587719298245614, 'recall': 0.7587719298245614, 'f1': 0.7587719298245615, 'support': None}, 'macro_avg': {'precision': 0.5781794186205951, 'recall': 0.5251997810618501, 'f1': 0.49699520529861696, 'support': None}, 'weighted_avg': {'precision': 0.7615654227738207, 'recall': 0.7587719298245614, 'f1': 0.7305164823821534, 'support': None}}
Loss: tensor(0.0195, 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(0.0003, 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(0.4078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, 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(5.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6064e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(8.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0291e-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(8.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, 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(0.1190, 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(0.0002, 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(2.0014e-05, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-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(0.0002, 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(0.0004, 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(0.0031, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6644e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, 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(0.0006, 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(8.0141e-05, device='cuda:0', grad_fn=<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(0.0074, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7741e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, 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(0.0002, 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(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4458, device='cuda:0', grad_fn=<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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5216e-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(0.0001, 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(0.0366, 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(0.0041, 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(0.0004, 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(0.0010, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-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(0.0018, 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(0.0003, 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(0.0005, 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(7.8294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3775e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_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(9.7045e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, 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(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, 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(3.7060e-05, device='cuda:0', grad_fn=<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(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, 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(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(0.0004, 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(7.5813e-05, 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(0.0030, 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(9.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7276e-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(5.7406e-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(3.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5586e-05, 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(0.0006, 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(7.7443e-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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2600e-05, device='cuda:0', grad_fn=<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(9.0435e-05, 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(0.0005, 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(9.5734e-05, 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(9.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0744, 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(0.0003, 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(0.0104, 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(0.0003, 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(0.0342, 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(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(0.0029, 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(0.9521, 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(0.0006, 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(0.0020, 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(0.0015, 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(0.0499, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5418, 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(0.0075, 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(0.0376, 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(0.0065, 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(0.0083, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9304e-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(0.0019, 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(0.0056, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, 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(0.0025, 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(0.0003, 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(0.0074, 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(9.7883e-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(0.0015, 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(0.0055, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, 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(0.0031, 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(0.0185, 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(8.3244e-05, 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(0.0002, 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(0.0013, 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(0.0057, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6713e-05, 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(9.3070e-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(0.0013, 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(0.0002, 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(0.0003, 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(0.0002, 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(0.0001, 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(0.2462, 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(0.0062, 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(0.3087, device='cuda:0', grad_fn=<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(0.0008, 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(0.0001, 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(9.3735e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, 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(0.0002, 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(0.0006, 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(0.1208, 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(0.0005, 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(0.0038, 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(0.0038, 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(0.0033, 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(0.0004, 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(0.0002, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9557e-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(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, 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(0.0045, 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(0.0006, 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(0.0005, 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(0.1034, 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(0.0005, 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(0.4995, device='cuda:0', grad_fn=<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(0.0044, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0001, 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(1.2574, device='cuda:0', grad_fn=<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(8.4593e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8006, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0003, 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(0.0002, 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(0.0001, 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(0.0089, 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(0.0003, 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(0.0018, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-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(0.0014, 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(0.0002, 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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-05, 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(0.0010, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9709, 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(0.0007, 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(0.0005, 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(0.0621, 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(0.0060, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, 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(0.1075, 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(0.0020, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, 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(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9800e-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(0.0003, 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(0.0017, 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(0.0010, 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(0.0016, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2412e-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(0.0029, 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(0.0016, 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(0.0085, 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(0.0114, 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(0.0005, 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(0.0037, 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(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5480e-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(0.0002, 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(0.0075, 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(0.0001, 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(0.0053, 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(9.4491e-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(9.9184e-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(0.0002, 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(0.0048, 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(0.0052, 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(0.4794, 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(0.0027, 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(0.0039, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1838, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0102e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0001, 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(0.0026, 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(0.0021, 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(0.0003, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3594e-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(0.0003, 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(0.0003, 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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0010, 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(0.0003, 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(0.0013, 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(0.0022, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, 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(0.0005, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, 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(0.0163, 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(0.0008, 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(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(0.0029, 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(0.0001, 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(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(0.2366, 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(0.0071, device='cuda:0', grad_fn=<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(0.0009, 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(0.0008, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1704e-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(0.0023, 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(0.0044, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0007, 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(6.4036e-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(5.4409e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1877e-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(2.0344, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, 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(0.0007, 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(0.0008, 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(0.0015, 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(0.0020, 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(0.0006, 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(0.0004, 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(0.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5577e-05, 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(0.0001, 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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.0139e-05, 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(9.9913e-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(0.0001, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-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(0.0002, 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(8.4048e-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(0.0001, 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(0.0016, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.0003, 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(0.0003, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(9.2643, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6042e-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(0.0006, 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(0.0009, 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(0.0021, 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(0.0004, 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(0.0035, 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(0.0057, 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(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0007, 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(0.0090, 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(0.0046, 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(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0018, 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(0.0011, 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(0.0009, 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(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, 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(0.0008, 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(0.0032, 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(0.0298, 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(0.0194, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0036, 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(0.3315, 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(0.0083, 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(0.0018, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, 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(0.0023, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0046, 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(0.0207, 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(0.0048, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0009, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, 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(0.0169, 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(0.0016, 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(0.0017, 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(0.0009, 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(0.0023, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, 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(0.0070, 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(0.0010, 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(0.0321, 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(0.0025, 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(0.0075, 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(0.0031, 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(0.0029, 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(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4890, 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(0.0037, 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(0.0010, 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(0.0022, 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(0.0053, 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(0.0024, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2968, 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(0.0049, 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(0.0008, 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(0.0044, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, 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(0.0050, 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(0.0024, 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(0.0092, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7592, 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(0.0002, 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(0.0018, 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(0.0016, 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(0.0033, 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(0.0078, 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(0.0098, 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(0.0028, 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(0.0017, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0061, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, 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(0.0012, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, 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(0.0109, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, 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(0.0010, 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(0.0006, 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(0.0011, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5383, 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(0.0009, 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(0.0072, 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(0.0014, 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(0.0007, 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(0.0009, 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(0.1745, 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(0.0012, 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(0.0078, 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(0.0007, 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(0.0009, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0048, 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(0.0007, 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
				 {'precision': {'support': 0.891156462585034, 'agreement': 0.8245614035087719, 'direct_attack': 0.4444444444444444, 'undercutter_attack': 0.3333333333333333, 'partial': 0.20512820512820512}, 'recall': {'support': 0.9703703703703703, 'agreement': 0.5222222222222223, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.4482758620689655, 'partial': 0.32}, 'f1': {'support': 0.9290780141843973, 'agreement': 0.6394557823129252, 'direct_attack': 0.34782608695652173, 'undercutter_attack': 0.38235294117647056, 'partial': 0.25}, 'support': {'support': 270, 'agreement': 90, 'direct_attack': 42, 'undercutter_attack': 29, 'partial': 25}, 'micro_avg': {'precision': 0.75, 'recall': 0.75, 'f1': 0.75, 'support': None}, 'macro_avg': {'precision': 0.5397247697999579, 'recall': 0.5093165480751687, 'f1': 0.509742564926063, 'support': None}, 'weighted_avg': {'precision': 0.7637813808668578, 'recall': 0.75, 'f1': 0.7463794192636888, '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
				 {'precision': {'support': 0.5833333333333334, 'agreement': 0.1282051282051282, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.05825242718446602}, 'recall': {'support': 0.725, 'agreement': 0.11627906976744186, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.15789473684210525}, 'f1': {'support': 0.6464968152866242, 'agreement': 0.12195121951219512, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.08510638297872342}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.43673469387755104, 'recall': 0.43673469387755104, 'f1': 0.43673469387755104, 'support': None}, 'macro_avg': {'precision': 0.1539581777445855, 'recall': 0.19983476132190942, 'f1': 0.17071088355550854, 'support': None}, 'weighted_avg': {'precision': 0.3491015226105379, 'recall': 0.43673469387755104, 'f1': 0.3867286801478993, 'support': None}}
Loss: tensor(1.9587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, 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
				 {'precision': {'support': 0.9327731092436975, 'agreement': 0.38235294117647056, 'direct_attack': 0.23786407766990292, 'undercutter_attack': 0.6666666666666666, 'partial': 0.0}, 'recall': {'support': 0.7928571428571428, 'agreement': 0.3023255813953488, 'direct_attack': 0.765625, 'undercutter_attack': 0.09230769230769231, 'partial': 0.0}, 'f1': {'support': 0.8571428571428571, 'agreement': 0.33766233766233766, 'direct_attack': 0.36296296296296293, 'undercutter_attack': 0.16216216216216217, 'partial': 0.0}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.5918367346938775, 'recall': 0.5918367346938775, 'f1': 0.5918367346938775, 'support': None}, 'macro_avg': {'precision': 0.4439313589513475, 'recall': 0.3906230833120368, 'f1': 0.34398606398606396, 'support': None}, 'weighted_avg': {'precision': 0.6860699619653686, 'recall': 0.5918367346938775, 'f1': 0.5883462258972463, 'support': None}}
Loss: tensor(0.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(1.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434, 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(1.3101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9884, 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(0.3120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(1.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8666, 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
				 {'precision': {'support': 0.9462809917355371, 'agreement': 0.2235294117647059, 'direct_attack': 0.4342105263157895, 'undercutter_attack': 0.49333333333333335, 'partial': 0.08333333333333333}, 'recall': {'support': 0.8178571428571428, 'agreement': 0.4418604651162791, 'direct_attack': 0.515625, 'undercutter_attack': 0.5692307692307692, 'partial': 0.02631578947368421}, 'f1': {'support': 0.8773946360153255, 'agreement': 0.29687500000000006, 'direct_attack': 0.4714285714285714, 'undercutter_attack': 0.5285714285714285, 'partial': 0.04}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.6510204081632653, 'recall': 0.6510204081632653, 'f1': 0.6510204081632653, 'support': None}, 'macro_avg': {'precision': 0.4361375192965398, 'recall': 0.474177833335575, 'f1': 0.4428539272030651, 'support': None}, 'weighted_avg': {'precision': 0.6889658151211768, 'recall': 0.6510204081632653, 'f1': 0.6622136622711481, 'support': None}}
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2876, 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(2.2922, 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(0.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, 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(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, 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(2.1141, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, 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(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(0.9603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8429, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1554, 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(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, 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(0.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0602, 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(1.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0061, 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(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4201, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7952, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2682, 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(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(1.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, 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(0.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1377, 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(1.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3555, 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(0.9006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2720, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(5.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9015, 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(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.5603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, 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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9344, 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(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, 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(1.1978, 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(0.5663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0261, 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(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2687, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4757, 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(0.7597, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, 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(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4702, 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(0.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, 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(0.0106, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, 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(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, 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(1.9773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(2.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, 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(2.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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
				 {'precision': {'support': 0.971830985915493, 'agreement': 0.25862068965517243, 'direct_attack': 0.296875, 'undercutter_attack': 0.4838709677419355, 'partial': 0.0}, 'recall': {'support': 0.7366548042704626, 'agreement': 0.6976744186046512, 'direct_attack': 0.59375, 'undercutter_attack': 0.23076923076923078, 'partial': 0.0}, 'f1': {'support': 0.8380566801619433, 'agreement': 0.3773584905660377, 'direct_attack': 0.3958333333333333, 'undercutter_attack': 0.3125, 'partial': 0.0}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 37}, 'micro_avg': {'precision': 0.5918367346938775, 'recall': 0.5918367346938775, 'f1': 0.5918367346938775, 'support': None}, 'macro_avg': {'precision': 0.4022395286625201, 'recall': 0.45176969072886897, 'f1': 0.38474970081226284, 'support': None}, 'weighted_avg': {'precision': 0.6829730808176566, 'recall': 0.5918367346938775, 'f1': 0.6068697460268959, 'support': None}}
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, 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(0.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2707, 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(0.5747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, 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(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, 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(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.6732, 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(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(3.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, 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(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.2274, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, 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(0.0042, 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(3.4306, 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(0.0482, 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(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, 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(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, 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(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, 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(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, 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(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, 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(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, 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(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, 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(1.4408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, 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(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, 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(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, 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(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, 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(0.1334, 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(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, 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(0.0116, 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(0.0113, 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(1.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, 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(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1883, 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(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, 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(0.0035, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5882, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, 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(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, 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(2.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1437, 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(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.6761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2509, 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(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, 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(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, 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(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, 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(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, 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(0.0083, 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(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, 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(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8350, 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(0.2802, 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(0.0152, 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(0.0265, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0058, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5579, 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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, 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(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, 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(0.0039, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0503, 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(0.0710, 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(0.0162, 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(0.0060, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1612, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, 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(4.2056, 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(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9383, 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(0.0167, 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(0.0016, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, 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(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, 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(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(1.3548, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, 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(2.8176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8367, 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(1.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7835, 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(0.0356, 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(2.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, 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(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, 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(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, 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>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, 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(0.3975, 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(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, 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(0.0045, 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(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, 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(0.0012, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, 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(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3306, 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(2.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, 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(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8144, 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(0.0204, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6705, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4371, 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(0.1016, 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(0.0524, 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(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, 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(0.2069, 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(1.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7997, 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(0.0009, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0472, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3386, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(1.8134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3999, 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(0.3172, 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(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0115, 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(0.0008, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0075, 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(0.0057, 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(2.9219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8719, 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(0.2680, 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(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4434, 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(0.1088, 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(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5673, 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(1.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, 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(0.1350, 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(0.0078, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(0.0009, 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(0.0104, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0007, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, 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(1.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6826, 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(0.0668, 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(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, 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(0.0659, 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(0.0006, 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(0.1088, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9175, 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(0.0005, 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(1.6071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, 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(0.0873, 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(0.0104, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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>)
			------------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
				 {'precision': {'support': 0.9566929133858267, 'agreement': 0.28888888888888886, 'direct_attack': 0.36, 'undercutter_attack': 0.4666666666666667, 'partial': 0.14285714285714285}, 'recall': {'support': 0.8647686832740213, 'agreement': 0.3023255813953488, 'direct_attack': 0.42857142857142855, 'undercutter_attack': 0.4307692307692308, 'partial': 0.21052631578947367}, 'f1': {'support': 0.9084112149532712, 'agreement': 0.2954545454545454, 'direct_attack': 0.391304347826087, 'undercutter_attack': 0.44800000000000006, 'partial': 0.1702127659574468}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 63, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.6510204081632653, 'recall': 0.6510204081632653, 'f1': 0.6510204081632653, 'support': None}, 'macro_avg': {'precision': 0.44302112235970503, 'recall': 0.44739224795990057, 'f1': 0.44267657483827005, 'support': None}, 'weighted_avg': {'precision': 0.6932547666235598, 'recall': 0.6510204081632653, 'f1': 0.669812971175186, 'support': None}}
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, 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(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, 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(0.0147, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, 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(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, 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(2.4545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3518, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(3.9381, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0591, 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(0.0048, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0040, 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(0.0007, 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(0.2582, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2131, 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(0.2045, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.1118, 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(0.0347, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0495, 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(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, 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(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.1217, 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(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, 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(0.0025, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0953, 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(0.0250, 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(0.0108, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.0113, 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(0.0022, 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(0.0114, 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(0.0033, 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(0.5699, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0010, 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(0.0057, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0106, 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(0.0036, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, 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(0.2830, 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(0.0205, 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(0.0074, 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(0.0115, 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(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, 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(0.0050, 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(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, 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(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, 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(0.0159, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, 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(0.0062, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3684, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, 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(0.5439, 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(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, 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(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, 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(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0025, 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(1.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1916, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3916, 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(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6293, 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(0.5151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, 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(0.0799, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, 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(0.7099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1691, 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(0.0014, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, 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(0.0055, 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(0.0127, device='cuda:0', grad_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(0.1107, 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(4.3300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, 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(0.0364, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, 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(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, 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(0.0918, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, 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(0.0004, 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(0.0048, 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(0.5973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, 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(0.0016, 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(0.0055, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, 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(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, 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(0.0062, 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(0.0046, 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(0.0003, 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(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5584, 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(0.0068, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, 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(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0028, 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(0.0845, 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(0.0162, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4257, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4034, 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(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2315, 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(0.0106, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, 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(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<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(0.0103, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.0206, 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(0.0299, 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(1.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.1159, 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(0.0015, 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(0.0454, 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(0.0679, 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(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249, 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(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.1181, device='cuda:0', grad_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(0.9951, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0800, 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(0.0231, 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(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6223, 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(0.4679, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, 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(0.0627, 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(0.0054, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(3.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, 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(0.0056, 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(0.0869, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4874, 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(0.0012, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0037, 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(0.0006, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, 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(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, 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(0.0402, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, 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(0.0369, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0026, 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(0.0085, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1296, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3350, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, 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(0.0104, 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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4638, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(0.0025, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3092, 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(0.0088, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, 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(0.0028, 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(0.0320, 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(0.0319, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, 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(0.0008, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0146, 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(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7903, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, 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(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, 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(0.0220, 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(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, 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(0.1061, 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(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, 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(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, 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(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1617, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3672, device='cuda:0', grad_fn=<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(0.0720, 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(0.0132, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, 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(0.0075, 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(0.0005, 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(0.0616, 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(0.0089, 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(2.6125, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3310, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2402, 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(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0009, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, 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(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, 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(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6338, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, 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(0.0005, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0408, 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(0.0034, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, 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(0.0423, device='cuda:0', grad_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(0.0033, 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(0.0044, 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(0.0082, 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(0.5360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, 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(0.0529, 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(1.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0007, 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(0.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, 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(0.1517, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, 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(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, 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(0.0006, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, 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(0.0068, 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(0.1110, 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(0.0038, 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(0.0065, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3870, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0013, 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(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, 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(3.3894, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, 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(0.0041, 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(0.0077, device='cuda: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 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
				 {'precision': {'support': 0.9323843416370107, 'agreement': 0.39622641509433965, 'direct_attack': 0.42, 'undercutter_attack': 0.42574257425742573, 'partial': 0.0}, 'recall': {'support': 0.9323843416370107, 'agreement': 0.4883720930232558, 'direct_attack': 0.328125, 'undercutter_attack': 0.671875, 'partial': 0.0}, 'f1': {'support': 0.9323843416370107, 'agreement': 0.4375, 'direct_attack': 0.368421052631579, 'undercutter_attack': 0.5212121212121212, 'partial': 0.0}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 64, 'partial': 38}, 'micro_avg': {'precision': 0.7081632653061225, 'recall': 0.7081632653061225, 'f1': 0.7081632653061225, 'support': None}, 'macro_avg': {'precision': 0.43487066619775516, 'recall': 0.4841512869320533, 'f1': 0.45190350309614213, 'support': None}, 'weighted_avg': {'precision': 0.6799291032684324, 'recall': 0.7081632653061225, 'f1': 0.6892837206652996, 'support': None}}
Loss: tensor(0.0738, 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(0.0182, 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(0.0009, 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(0.0026, 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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, 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(0.0074, 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(0.0006, 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(0.0003, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0040, 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(0.0009, 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(0.0031, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, 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(0.0089, 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(0.0004, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0003, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, 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(0.0356, 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(0.0365, 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(0.0129, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, 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(0.0006, 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(0.0003, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_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(0.0236, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, 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(0.0025, 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(0.0056, 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(0.0015, 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(0.0007, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, 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(0.0050, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0004, 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(0.0029, 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(0.0042, 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(0.0033, 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(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, 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(0.0053, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, 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(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, 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(0.0006, 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(0.0005, 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(0.0093, 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(0.1114, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4930, 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(0.0049, 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(0.0025, 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(0.0026, 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(0.0005, 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(0.0037, 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(0.0043, 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(0.0014, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, 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(0.0035, 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(0.0021, 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(0.0039, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, 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(0.0018, 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(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0006, 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(0.0006, 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(0.0098, 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(0.0079, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.0073, 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(0.0028, 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(0.0065, device='cuda:0', grad_fn=<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(0.0595, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0004, 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(0.0058, device='cuda:0', grad_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(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, 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(0.0035, 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(0.0004, 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(0.0007, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, 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(0.0026, 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(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, 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(0.0113, 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(0.0161, 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(0.0003, 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(0.0232, 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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, 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(0.0026, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, 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(0.0011, 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(0.0017, 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(0.0068, 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(1.3582, 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(0.0057, 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(0.0023, 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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, 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(0.0685, 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(0.0019, 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(0.0006, 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(0.0007, 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(0.0061, 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(0.0005, 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(0.0064, 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(0.0047, 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(0.0003, 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(0.0003, 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(0.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.0009, 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(0.0002, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0002, 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(0.0011, 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(0.8604, 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(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(0.0019, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, 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(0.0007, 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(0.0028, 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(0.0028, 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(0.0017, 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(0.0026, 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(0.0007, 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(0.0065, 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(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, 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(0.0052, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, 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(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0845, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, 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(0.0346, 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(0.0022, 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(0.0013, 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(0.0137, 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(0.0007, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0033, 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(0.0013, 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(0.0025, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, 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(0.0014, 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(0.0048, 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(0.0002, 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(0.0002, 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(0.0036, 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(0.0041, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, 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(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, 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(0.0016, 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(0.0020, 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(0.0017, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, 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(0.4720, 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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3199, 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(0.0009, 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(0.0002, 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(0.0014, 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(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.0005, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0020, 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(0.0016, 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(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(0.0037, 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(0.0093, device='cuda:0', grad_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(0.0754, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0005, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0014, 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(0.0059, 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(0.0069, device='cuda:0', grad_fn=<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(0.0037, device='cuda:0', grad_fn=<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(0.0017, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, 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(0.0484, 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(0.0032, 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(0.0002, 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(0.0008, 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(0.0005, 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(0.0806, 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(0.0081, 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(0.0002, 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(0.0018, 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(0.0023, 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(0.0011, 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(0.0009, 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(0.0016, 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(0.0026, 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(0.0018, 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(0.0004, 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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<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(0.0075, 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(0.0023, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9281, 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(0.0014, 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(0.0007, 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(0.0004, 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(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(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(0.0644, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0015, 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(0.0001, 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(0.0002, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9820e-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(0.0002, 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(0.0249, 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(0.0045, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, 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(0.0023, 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(0.0061, 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(0.0008, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0190, 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(0.0006, 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(0.0284, 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(0.0004, 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(0.0017, 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(0.0002, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, 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(0.0301, 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(0.0021, 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(0.0002, 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(0.0292, 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(0.0025, 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(0.0078, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4353, 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(0.0049, 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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, 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(0.0007, 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(0.0045, 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(0.0039, 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(0.0016, 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(0.0379, device='cuda:0', grad_fn=<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(0.0042, device='cuda:0', grad_fn=<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(0.0004, 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(0.0025, 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(0.0005, 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(0.0009, 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(0.0002, 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(0.0009, 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(0.0022, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0010, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1182, 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(0.0009, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0040, 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(0.0008, 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(0.0003, 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(0.0040, 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(0.0135, device='cuda:0', grad_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(0.0081, 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(0.0009, 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(0.0040, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0006, 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(0.0374, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0007, 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(0.0021, 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(0.0017, 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(0.0004, 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(0.0007, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9037, 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(0.0140, 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(0.0006, 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(0.0014, 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(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(0.0011, 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(0.0001, 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(0.0164, 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(0.0001, 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(0.0005, 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(0.0020, 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(0.0022, 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(0.0005, 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(0.0051, 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(0.0022, 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(0.0008, 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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-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(0.0002, 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(0.0018, 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(0.0122, device='cuda:0', grad_fn=<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(0.0026, 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(0.0021, 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(0.0005, 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(0.0004, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0002, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0033, 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(0.0017, 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(0.0016, 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(0.0028, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0001, 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(0.0031, 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(0.0015, 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(0.0031, 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(0.0002, 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(0.0072, 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(0.0015, 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(0.0030, 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(0.0037, 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(0.0009, 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(0.0010, 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(0.0002, 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(0.0066, 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(0.0707, 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(0.0009, 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(0.0012, 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(0.0025, 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(0.0002, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, 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(0.1381, 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(0.0107, device='cuda:0', grad_fn=<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(0.0003, 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(0.0018, 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(0.0030, 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(0.0047, 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(0.0005, 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(0.0031, 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(0.0010, 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(0.0003, 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(0.0078, 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(0.0016, 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(0.0042, 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(8.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0998, device='cuda:0', grad_fn=<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(0.0004, 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(0.0004, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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
				 {'precision': {'support': 0.9487179487179487, 'agreement': 0.4318181818181818, 'direct_attack': 0.3472222222222222, 'undercutter_attack': 0.4105263157894737, 'partial': 0.16666666666666666}, 'recall': {'support': 0.9217081850533808, 'agreement': 0.4418604651162791, 'direct_attack': 0.390625, 'undercutter_attack': 0.6, 'partial': 0.02702702702702703}, 'f1': {'support': 0.9350180505415162, 'agreement': 0.4367816091954023, 'direct_attack': 0.3676470588235294, 'undercutter_attack': 0.4875, 'partial': 0.046511627906976744}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 37}, 'micro_avg': {'precision': 0.7, 'recall': 0.7, 'f1': 0.7, 'support': None}, 'macro_avg': {'precision': 0.4609902670428986, 'recall': 0.4762441354393374, 'f1': 0.4546916692934849, 'support': None}, 'weighted_avg': {'precision': 0.6943490302512859, 'recall': 0.7, 'f1': 0.6907337212139436, 'support': None}}
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(0.0003, 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(0.0213, 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(0.0023, 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(0.0027, 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(0.0009, 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(0.1109, 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(0.0011, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, 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(0.0056, 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(0.0004, 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(0.0063, 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(0.0014, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9366e-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(0.0038, 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(0.0006, 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(0.0039, device='cuda:0', grad_fn=<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(0.0006, 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(0.0006, 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(0.0002, 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(0.0003, 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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7842e-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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0284e-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(0.0002, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0002, 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(0.0057, 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(0.0066, 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(0.0003, 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(0.0003, 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(0.0004, 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(0.0026, 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(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(0.0012, 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(0.0020, 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(0.0031, 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(0.0098, 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(0.0017, 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(0.0014, 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(0.0008, 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(0.0005, 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(0.1581, 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(0.0037, 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(0.0034, 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(0.0208, 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(0.0014, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8004e-05, 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(0.0042, 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(0.0042, 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(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(0.0002, 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(0.0116, 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(0.0053, 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(0.0018, 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(0.0007, 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(0.0009, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8010, 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(2.1349, 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(0.0041, 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(7.6812e-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(0.0017, 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(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(0.0010, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9124e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0004, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, 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(0.0003, 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(0.0163, device='cuda:0', grad_fn=<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(0.0092, 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(0.0012, 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(0.0010, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7882e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2555e-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(0.0002, 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(0.0024, device='cuda:0', grad_fn=<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(0.0001, 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(0.0001, 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(0.0124, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-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(0.0273, 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(0.0092, 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(0.0064, 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(0.0009, 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(9.9064e-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(0.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, 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(9.6430e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.4095, 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(0.0001, 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(0.0001, 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(0.0046, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, 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(0.0058, 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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9255e-05, 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(0.0012, 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(0.0046, device='cuda:0', grad_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(0.0001, 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(0.0016, 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(0.0003, 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(0.0001, 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(0.0009, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8246e-05, 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(0.0001, 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(0.0001, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0006, 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(0.0038, 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(0.0004, 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(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, 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(0.0002, device='cuda:0', grad_fn=<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(0.0023, 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(0.0295, 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(0.0016, 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(0.0045, 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(5.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0033, 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(0.0014, 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(0.0043, 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(0.0158, 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(0.0031, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0055, 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(0.0012, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, 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(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, 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(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, 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(0.0025, 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(0.0010, 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(0.0128, 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(0.0053, 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(0.0028, 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(0.0039, 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(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, 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(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0009, 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(0.0017, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.0018, 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(0.0023, 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(0.0008, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0030, 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(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0010, 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(0.0009, 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(0.1305, 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(0.0041, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2041, 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(0.0138, 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(0.0040, 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(0.0003, 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(0.0019, 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(0.0012, 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(0.0011, 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(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(0.0016, 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(0.0007, 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(0.0011, 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(0.0002, 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(0.0045, 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(0.0038, device='cuda:0', grad_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(0.0018, 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(0.0016, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, 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(0.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0130, 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(0.0006, 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(0.0061, 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(0.0002, 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(0.0030, 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(0.0060, 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(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(0.0002, 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(0.0007, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, 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(0.0026, 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(0.0025, 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(0.0016, 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(0.0003, 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(0.0107, 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(0.0032, 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(0.0001, 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(0.0005, 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(0.0016, 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(0.0021, 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(0.0001, 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(0.0029, 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(0.0003, 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(0.0006, 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(0.0003, 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(0.0002, 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(0.0004, 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(0.0011, 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(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(0.0024, 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(0.0087, 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(0.0002, 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(0.0011, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, 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(0.0002, 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(0.0002, 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(0.0012, 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(0.0001, 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(0.0018, 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(0.0059, 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(0.0010, 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(0.0032, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1715, 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(0.0008, 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(0.0025, 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(0.0009, 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(0.0020, device='cuda:0', grad_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(0.0003, 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(0.0081, 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(0.0039, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7489e-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(9.2494e-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(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(0.0011, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1162, 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(0.0029, 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(0.0002, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8316e-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(0.0004, 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(0.0003, 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(0.0003, 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(0.0002, 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(0.0015, 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(0.0053, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0047, device='cuda:0', grad_fn=<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(0.0010, 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(0.0002, 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(0.0084, 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(0.0002, 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(0.0060, device='cuda:0', grad_fn=<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(0.0007, 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(0.0010, 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(0.0016, 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(0.0002, 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(0.0045, 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(0.0009, 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(0.0002, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5461e-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(0.0010, 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(7.5662e-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(0.0019, 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(0.0003, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0014, 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(0.0031, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4885e-05, 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(0.0043, 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(0.0393, 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(0.0038, 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(0.0123, 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(0.0002, 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(0.0016, 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(0.0033, 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(0.7229, 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(0.0010, 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(0.0003, 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(0.0013, 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(0.0007, 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(0.0001, 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(0.0030, 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(0.0014, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0009, 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(0.0009, 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(7.9295e-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(0.0002, 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(0.0039, 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(0.0003, 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(0.0079, 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(0.0149, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, 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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0017, 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(0.0002, 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(0.0011, 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(0.0089, device='cuda:0', grad_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(0.0118, 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(0.0068, 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(0.0003, 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(0.0074, 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(0.0003, 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(0.0013, 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(0.0002, 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(0.0034, 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(0.0012, 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(0.0043, 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(0.0002, 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(0.0007, 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(0.0004, 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(0.0076, device='cuda:0', grad_fn=<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(0.0014, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, 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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<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(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(0.0284, 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(0.0040, device='cuda:0', grad_fn=<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(0.0027, 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(0.0001, 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(0.0019, 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(0.0002, 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(0.0026, 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(0.0014, 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(0.0004, 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(0.0012, 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(0.0002, 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(0.0005, 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(0.0074, 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(0.0050, 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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1737e-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(0.0002, 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(0.0079, 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(0.0007, 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(0.0002, 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(0.0009, 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(0.0015, 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(0.0005, 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(0.0013, 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(0.0026, 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(0.0007, 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(0.0013, 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(0.0035, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3372e-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(0.0002, 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(0.0023, 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(0.0027, 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(0.0049, device='cuda:0', grad_fn=<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(0.0006, 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(0.0004, 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(0.0032, 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(0.0006, 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(0.0013, 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(0.0001, 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(0.0004, 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(0.0020, 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(0.0016, 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(0.0022, 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(0.0018, 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(0.0002, 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(0.0020, 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(0.0002, 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(0.0002, 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(0.0008, 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(0.0062, device='cuda:0', grad_fn=<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(0.0008, 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(0.8898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, 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(0.0007, 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>)
			------------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
				 {'precision': {'support': 0.9257950530035336, 'agreement': 0.37777777777777777, 'direct_attack': 0.4426229508196721, 'undercutter_attack': 0.4787234042553192, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9323843416370107, 'agreement': 0.3953488372093023, 'direct_attack': 0.421875, 'undercutter_attack': 0.6923076923076923, 'partial': 0.02702702702702703}, 'f1': {'support': 0.9290780141843972, 'agreement': 0.38636363636363635, 'direct_attack': 0.432, 'undercutter_attack': 0.5660377358490566, 'partial': 0.045454545454545456}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 37}, 'micro_avg': {'precision': 0.7183673469387755, 'recall': 0.7183673469387755, 'f1': 0.7183673469387754, 'support': None}, 'macro_avg': {'precision': 0.4735552657426891, 'recall': 0.4937885796362065, 'f1': 0.47178678637032706, 'support': None}, 'weighted_avg': {'precision': 0.6961703239861355, 'recall': 0.7183673469387755, 'f1': 0.7016465905335896, 'support': None}}
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(9.7428e-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(9.2312e-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(0.0039, 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(0.0032, 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(0.0179, device='cuda:0', grad_fn=<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(0.0004, 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(0.0035, 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(0.0002, 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(0.0010, 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(0.0017, 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(0.0042, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1525e-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(0.0031, 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(0.0034, 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(0.0002, 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(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(0.0017, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8822e-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(0.0037, 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(0.0011, 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(0.0010, 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(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(8.0656e-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(0.0033, 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(0.0001, 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(0.0001, 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(0.0002, 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(0.0005, 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(9.2857e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0688e-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(0.0002, device='cuda:0', grad_fn=<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(9.8426e-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(0.0017, 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(0.0049, 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(0.0047, 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(0.0012, 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(0.0004, 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(0.0001, 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(0.0037, device='cuda:0', grad_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(0.0024, 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(0.0033, 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(0.0002, 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(0.0062, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0969e-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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8246e-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(0.0016, 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(0.0001, 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(0.0012, 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(0.0007, 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(7.4875e-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(8.4260e-05, device='cuda:0', grad_fn=<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(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(0.0002, 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(0.0002, 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(0.0011, 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(0.0008, 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(0.0008, 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(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(0.0013, 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(0.0013, 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(0.0015, 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(0.0023, 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(0.0012, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2059e-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(0.0001, 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(0.0002, 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(8.5077e-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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5662e-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(0.0005, 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(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(0.0007, 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(0.0006, 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(0.0003, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4976e-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(0.0008, 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(0.0015, 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(0.0001, 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(7.1241e-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(0.0030, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9406e-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(0.0007, 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(9.3342e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, 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(0.0439, 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(0.0003, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4977e-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(9.4432e-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(0.0019, 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(0.0002, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6560e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0005, 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(0.0013, 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(0.0019, 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(0.0002, 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(6.8124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2595e-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(9.0647e-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(0.0006, 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(0.0002, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0002, 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(0.0007, 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(0.0033, 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(0.0014, 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(0.0001, 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(0.0004, 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(0.0035, 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(0.0050, 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(0.0036, 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(0.0009, 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(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(0.0008, 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(0.0008, 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(0.0005, 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(0.0003, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0001, 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(0.0003, 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(0.0037, 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(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(0.0068, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9819e-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(0.0001, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2616e-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(0.0030, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0012e-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(0.0028, 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(0.0003, 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(0.0009, 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(0.0002, 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(0.0008, 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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-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(0.0013, 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(7.9749e-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(0.0007, 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(7.8387e-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(0.0007, 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(0.0002, 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(0.0158, 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(7.7478e-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(0.0019, 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(0.0008, 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(0.0006, 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(0.0003, 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(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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4432e-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(0.0001, 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(0.0016, 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(0.0025, 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(0.0002, 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(0.0024, 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(0.0002, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, 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(0.0004, 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(0.0002, 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(0.0002, 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(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(0.0008, 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(0.0016, 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(0.0017, 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(0.0006, 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(6.2553e-05, 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(5.9313e-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(7.2090e-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(0.0020, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2846e-05, 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(0.0018, 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(0.0019, 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(0.0001, 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(0.0002, 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(0.0001, 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(0.0004, 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(0.0004, 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(0.0011, 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(0.0008, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4099e-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(0.0001, 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(0.0016, 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(0.0004, 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(0.0007, 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(8.0082e-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(0.0007, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<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(6.8790e-05, 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(0.0050, device='cuda:0', grad_fn=<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(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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-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(2.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7035e-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(8.2776e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9042e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, 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(0.0032, 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(0.0012, 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(0.0017, 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(0.0002, 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(0.0007, 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(0.0005, 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(0.0016, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, 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(5.6255e-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(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(0.0041, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7961e-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(0.0001, 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(9.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<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(0.0006, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8245e-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(0.0002, 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(0.0008, 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(0.0001, 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(0.0001, 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(0.0008, 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(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-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(0.0030, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7035e-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(0.0090, device='cuda:0', grad_fn=<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(0.0001, 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(8.6681e-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(6.0766e-05, 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(0.0686, 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(0.0007, 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(7.1998e-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(0.0006, 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(0.0002, 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(0.0001, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3675e-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(0.0003, 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(0.0013, 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(0.0009, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0496e-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(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7378e-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(0.0005, 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(0.0006, 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(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0213e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, 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(0.0001, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0173e-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(0.0022, 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(0.0003, 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(0.0001, 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(6.6216e-05, 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(0.0013, 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(0.0028, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5247e-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(0.0002, 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(9.7822e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5167e-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(0.0017, 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(0.0001, 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(0.0103, 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(9.5794e-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(0.0003, 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(0.0002, 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(0.0024, 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(0.0001, 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(0.0013, 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(0.0010, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6409e-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(8.1656e-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(0.0002, 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(0.0002, 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(0.0001, 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(0.0013, 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(0.0011, 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(6.1826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1919e-05, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-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(0.0004, 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(0.0001, 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(0.0007, 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(0.0007, 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(0.0053, 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(0.1837, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1747e-05, 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(0.0014, 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(0.0011, 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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2997e-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(0.0018, 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(0.0019, 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(0.0012, 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(0.0011, 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(0.0060, 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(8.6197e-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(7.0061e-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(0.0002, 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(0.0042, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0007, 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(0.0002, 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(0.0003, 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(8.9709e-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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0012, 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(0.0001, 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(0.0009, 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(8.0536e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, 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(0.0088, 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(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(8.2020e-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(0.0028, 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(0.0008, 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(0.0003, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7105e-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(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7852e-05, 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(0.0007, 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(4.6931e-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(0.0002, 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(0.0001, 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(0.0028, 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(0.0125, 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(0.0012, 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(0.0029, 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(0.0018, 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(0.0033, 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(0.0028, 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(0.0007, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5854e-05, device='cuda:0', grad_fn=<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(0.0012, 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(0.0011, 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(0.0007, 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(0.0009, 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(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7610e-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(0.0056, 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>)
			------------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
				 {'precision': {'support': 0.9452554744525548, 'agreement': 0.375, 'direct_attack': 0.45098039215686275, 'undercutter_attack': 0.44954128440366975, 'partial': 0.125}, 'recall': {'support': 0.9217081850533808, 'agreement': 0.42857142857142855, 'direct_attack': 0.359375, 'undercutter_attack': 0.7538461538461538, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9333333333333335, 'agreement': 0.39999999999999997, 'direct_attack': 0.4, 'undercutter_attack': 0.5632183908045978, 'partial': 0.043478260869565216}, 'support': {'support': 281, 'agreement': 42, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7142857142857143, 'recall': 0.7142857142857143, 'f1': 0.7142857142857143, 'support': None}, 'macro_avg': {'precision': 0.46915543020261746, 'recall': 0.4979633113889294, 'f1': 0.46800599700149925, 'support': None}, 'weighted_avg': {'precision': 0.7024484018478482, 'recall': 0.7142857142857143, 'f1': 0.6998531346571614, 'support': None}}
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(0.0013, 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(0.0011, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0182e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-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(0.0006, 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(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8528e-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(0.0016, 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(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8123e-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(0.0002, device='cuda:0', grad_fn=<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(0.0008, 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(0.0012, 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(0.0004, 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(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6400e-05, device='cuda:0', grad_fn=<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(0.0032, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6671e-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(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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6469e-05, device='cuda:0', grad_fn=<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(7.4542e-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(7.3664e-05, device='cuda:0', grad_fn=<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(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9940e-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(0.0008, 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(0.0030, device='cuda:0', grad_fn=<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(0.0004, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-05, device='cuda:0', grad_fn=<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(6.6367e-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(0.0003, 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(0.0013, 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(0.0017, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6801e-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(6.1372e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-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(0.0009, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7459e-05, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2685e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7508e-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(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(0.0022, 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(6.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-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(0.0007, 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(7.5328e-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(9.9306e-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(0.0007, 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(0.0003, 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(0.0020, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5429e-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(0.0005, 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(0.0008, 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(0.0002, 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(0.0010, 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(0.0013, 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(0.0003, 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(0.0005, 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(0.0001, 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(0.0001, 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(0.0008, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6610e-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(0.0007, 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(0.0007, 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(0.0008, 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(0.0008, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6509e-05, device='cuda:0', grad_fn=<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(0.0015, 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(0.0006, 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(7.6994e-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(0.0003, 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(0.0004, 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(9.7971e-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(0.0013, 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(0.0004, 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(0.0086, 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(0.0001, 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(0.0011, 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(8.4563e-05, device='cuda:0', grad_fn=<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(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0014, 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(0.0001, 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(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-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(0.0014, 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(0.0004, 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(0.0008, 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(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(0.0027, 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(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(0.0013, 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(0.0004, 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(0.0021, 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(0.0009, 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(0.0003, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1384e-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(0.0004, 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(0.0019, 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(0.0001, 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(0.0001, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(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(0.0007, 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(0.0003, 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(0.0005, 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(0.0004, 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(0.0011, 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(5.6104e-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(0.0005, 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(0.0019, 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(0.0018, 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(0.0029, 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(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(0.0004, 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(0.0003, 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(0.0123, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7599, device='cuda:0', grad_fn=<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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-05, device='cuda:0', grad_fn=<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(6.0524e-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(0.0006, 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(0.0004, 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(0.0001, 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(0.0003, 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(0.0018, 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(0.0002, 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(0.0011, 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(0.0008, 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(6.7912e-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(9.0102e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7620e-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(0.0005, 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(9.5158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-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(5.9768e-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(0.0012, 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(0.0004, 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(9.8095e-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(5.7678e-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(0.0009, 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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9174e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7115e-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(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(9.8063e-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(0.0016, 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(0.0003, 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(0.0026, 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(0.0008, 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(0.0005, 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(0.0008, 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(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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-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(0.0004, 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(0.0002, 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(7.9900e-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(0.0015, 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(9.7853e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-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(7.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4663e-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(0.0004, 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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-05, 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(9.1526e-05, device='cuda:0', grad_fn=<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(0.0002, 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(8.6924e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5885e-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(4.7657e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6833e-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(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3503e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5683e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-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(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(0.0002, device='cuda:0', grad_fn=<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(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6216e-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(0.0014, 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(0.0007, 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(0.0005, 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(9.4099e-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(0.0011, 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(8.5985e-05, 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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5622e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2089e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-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(0.0008, 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(8.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-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(0.0006, 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(6.4460e-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(9.6338e-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(0.0002, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1635e-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(0.0005, 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(0.0001, 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(8.4047e-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(0.0004, 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(0.0005, 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(0.0005, 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(8.5802e-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(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(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-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(5.4530e-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(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5438e-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(1.2529, device='cuda:0', grad_fn=<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(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6851e-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(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6871e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8567e-05, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2171e-05, device='cuda:0', grad_fn=<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(0.0040, 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(0.0006, 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(8.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-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(7.2998e-05, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-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(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-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(9.2918e-05, device='cuda:0', grad_fn=<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(6.1947e-05, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4108e-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(7.0969e-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(7.2301e-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(0.0036, 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(0.0002, 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(0.0003, 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(0.0004, 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(0.0007, 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(8.5076e-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(0.0015, 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(6.2158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(8.0233e-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(0.0014, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3471e-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(6.1977e-05, device='cuda:0', grad_fn=<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(0.0047, 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(9.4762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6519e-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(8.7741e-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(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4894e-05, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6782e-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(0.0001, 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(0.0011, 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(6.8305e-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(6.4430e-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(0.0002, device='cuda:0', grad_fn=<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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7539e-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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3239e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1463e-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(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, 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(0.0008, 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(0.0039, 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(0.0005, 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(6.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8174e-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(4.5568e-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(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8968e-05, device='cuda:0', grad_fn=<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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-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(7.4723e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3129e-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(5.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2785e-05, 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(0.0020, 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(0.0003, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8507e-05, 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(0.0003, 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(0.0004, 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(0.0012, 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(5.2047e-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(6.5762e-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(0.0003, 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(0.0007, 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(9.7307e-05, device='cuda:0', grad_fn=<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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0009, 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(0.0001, 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(8.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5086e-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(0.0007, 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(0.0004, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-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(0.0011, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2857e-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(4.5296e-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(8.0627e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<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(4.9201e-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(0.0015, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3886e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-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(0.0016, 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(6.8819e-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(5.5256e-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(4.6870e-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(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8980e-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(0.0006, 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(0.0007, 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(4.2147e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-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(0.0027, 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(9.9148e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5005e-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(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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
				 {'precision': {'support': 0.9350180505415162, 'agreement': 0.3584905660377358, 'direct_attack': 0.4807692307692308, 'undercutter_attack': 0.47058823529411764, 'partial': 0.16666666666666666}, 'recall': {'support': 0.9217081850533808, 'agreement': 0.4523809523809524, 'direct_attack': 0.390625, 'undercutter_attack': 0.7384615384615385, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9283154121863799, 'agreement': 0.4, 'direct_attack': 0.43103448275862066, 'undercutter_attack': 0.5748502994011976, 'partial': 0.045454545454545456}, 'support': {'support': 281, 'agreement': 42, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7183673469387755, 'recall': 0.7183673469387755, 'f1': 0.7183673469387754, 'support': None}, 'macro_avg': {'precision': 0.4823065498618534, 'recall': 0.5058982930739111, 'f1': 0.47593094796014873, 'support': None}, 'weighted_avg': {'precision': 0.7050764803519035, 'recall': 0.7183673469387755, 'f1': 0.7027252651209696, 'support': None}}
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1826e-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(6.5429e-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(0.0006, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1011e-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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9514e-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(0.0003, 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(0.0020, 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(9.3764e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0021, 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(5.1502e-05, device='cuda:0', grad_fn=<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(6.8365e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-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(7.2543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7367e-05, device='cuda:0', grad_fn=<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(6.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(5.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1442e-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(0.0009, 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(0.0002, 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(0.0007, 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(0.0010, 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(0.0002, 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(8.8710e-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(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-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(5.8556e-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(5.7527e-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(0.0032, 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(0.0011, 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(0.0008, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0014, 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(8.7862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(0.0008, 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(0.0001, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-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(0.0001, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4399e-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(0.0006, 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(0.0013, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-05, device='cuda:0', grad_fn=<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(5.5922e-05, device='cuda:0', grad_fn=<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(6.8668e-05, device='cuda:0', grad_fn=<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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6812e-05, device='cuda:0', grad_fn=<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(6.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-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(0.0013, 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(0.0012, 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(7.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-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(0.0008, 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(0.0007, 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(5.5045e-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(9.8031e-05, device='cuda:0', grad_fn=<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(0.0012, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9891e-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(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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7467e-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(6.3067e-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(0.0004, 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(0.0004, 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(0.0004, 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(7.8992e-05, 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(0.0002, 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(7.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-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(4.5568e-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(8.2958e-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(0.0015, 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(0.0014, 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(8.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0312e-05, device='cuda:0', grad_fn=<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(0.0015, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2887e-05, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1817e-05, 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(0.0010, 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(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(0.0001, 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(0.0011, 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(6.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6580e-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(0.0009, 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(0.0003, 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(0.0004, 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(4.7535e-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(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3037e-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(0.0004, 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(4.7021e-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(0.0001, 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(0.0012, 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(0.0012, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-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(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(9.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7306e-05, 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(4.8262e-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(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(9.2191e-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(0.0001, 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(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7781e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-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(0.0003, 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(7.1090e-05, device='cuda:0', grad_fn=<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(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0005, 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(0.0010, 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(0.0007, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0505e-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(0.0015, 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(8.6530e-05, device='cuda:0', grad_fn=<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(0.0006, 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(7.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(6.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-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(6.8122e-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(6.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-05, 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(0.0020, device='cuda:0', grad_fn=<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(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-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(6.5338e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(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(0.0019, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(4.0815e-05, device='cuda:0', grad_fn=<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(7.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-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(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-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(0.0006, 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(6.9788e-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(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8416e-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(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-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(0.0005, 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(5.9949e-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(4.4569e-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(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(4.3630e-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(0.0020, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(5.8617e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6137e-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(6.0464e-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(0.0004, 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(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-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(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4609e-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(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5976e-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(0.0022, 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(7.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1442e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8325e-05, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2482e-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(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(0.0015, 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(0.0004, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9063e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6734e-05, device='cuda:0', grad_fn=<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(0.0006, 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(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, 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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0131e-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(0.0010, 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(0.0006, 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(8.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7285e-05, device='cuda:0', grad_fn=<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(8.3442e-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(7.8548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2522e-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(9.1462e-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(7.4905e-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(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4621e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-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(4.9292e-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(8.7296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7588e-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(0.0007, 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(0.0020, device='cuda:0', grad_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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-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(0.0006, 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(5.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0001, 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(8.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4552e-05, device='cuda:0', grad_fn=<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(0.0002, 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(4.6688e-05, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9775e-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(4.8747e-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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0506e-05, device='cuda:0', grad_fn=<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(5.0685e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, 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(0.0025, 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(0.0009, 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(0.0008, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5552e-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(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-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(7.3178e-05, 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(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1332e-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(7.6296e-05, 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(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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9343e-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(0.0015, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6516e-05, device='cuda:0', grad_fn=<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(3.2277e-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(0.0008, 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(0.0002, 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(0.0013, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(9.0486e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8685e-05, device='cuda:0', grad_fn=<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(4.5598e-05, device='cuda:0', grad_fn=<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(4.9927e-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(8.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<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(6.5640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7636e-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(7.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-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(4.0633e-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(0.0003, 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(0.0002, 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(3.3215e-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(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(0.0003, 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(6.7972e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-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(5.3773e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0015, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-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(0.0003, 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(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4885e-05, device='cuda:0', grad_fn=<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(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(6.1523e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6742e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4813e-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(8.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5993e-05, device='cuda:0', grad_fn=<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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6961e-05, device='cuda:0', grad_fn=<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(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-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(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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<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(8.9948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7811e-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(0.0011, 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(0.0005, 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(0.0009, 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(0.0008, 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(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(7.1302e-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(0.0003, 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(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-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(0.0003, 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(0.0009, 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(8.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3469e-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(0.0001, 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(7.6963e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-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(5.9978e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-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(0.0004, 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(0.0015, 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(0.0006, 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(0.0002, 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(0.0002, 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(6.2523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8375e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<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(0.0013, 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(6.9636e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0003, 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(0.0008, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.5962e-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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-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(0.0003, 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(0.0002, 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(7.4148e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4551e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9347826086956522, 'agreement': 0.375, 'direct_attack': 0.43283582089552236, 'undercutter_attack': 0.46153846153846156, 'partial': 0.125}, 'recall': {'support': 0.9214285714285714, 'agreement': 0.4186046511627907, 'direct_attack': 0.453125, 'undercutter_attack': 0.6461538461538462, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9280575539568345, 'agreement': 0.3956043956043956, 'direct_attack': 0.44274809160305345, 'undercutter_attack': 0.5384615384615385, 'partial': 0.043478260869565216}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.710204081632653, 'recall': 0.710204081632653, 'f1': 0.710204081632653, 'support': None}, 'macro_avg': {'precision': 0.46583137822592724, 'recall': 0.49312557164377846, 'f1': 0.4696699680990776, 'support': None}, 'weighted_avg': {'precision': 0.6945216795348899, 'recall': 0.710204081632653, 'f1': 0.6976635834582482, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5692e-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(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, 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(4.7082e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, 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(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(9.5324e-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(9.2615e-05, 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(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, 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(6.7184e-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(8.9889e-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(0.0006, 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(6.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-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(4.1692e-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(0.0001, 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(3.6636e-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(0.0001, 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(0.0008, 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(0.0003, 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(4.0421e-05, device='cuda:0', grad_fn=<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(0.0001, 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(7.7456e-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(9.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-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(0.0003, 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(8.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-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(8.3079e-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(4.2147e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9594e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4986e-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(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(0.0007, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-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(9.0315e-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(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-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(0.0002, device='cuda:0', grad_fn=<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(6.2280e-05, device='cuda:0', grad_fn=<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(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3974e-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(8.8586e-05, 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(0.0009, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6971e-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(0.0004, 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(5.9011e-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(0.0017, 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(9.9722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4513e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<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(7.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8841e-05, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6557e-05, device='cuda:0', grad_fn=<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(0.0017, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, 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(4.6809e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1625e-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(0.0012, 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(9.5865e-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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2101e-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(8.1202e-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(0.0001, 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(0.0009, 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(0.0013, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-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(4.0481e-05, device='cuda:0', grad_fn=<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(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399e-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(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2877e-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(5.0957e-05, 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(0.0006, 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(0.0014, 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(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8941e-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(0.0026, 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(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(0.0009, 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(0.0004, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3641e-05, device='cuda:0', grad_fn=<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(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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1141e-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(7.0031e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-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(6.2886e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-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(6.4611e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7084e-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(0.0011, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-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(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8132e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-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(4.3570e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-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(9.0768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8185e-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(0.0002, 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(0.0012, 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(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, 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(0.0002, 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(0.0001, 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(0.0003, 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(5.9798e-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(3.2943e-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(0.0003, 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(0.0003, 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(0.0004, 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(7.6205e-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(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1371e-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(0.0001, 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(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(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0290e-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(0.0007, 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(0.0001, 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(9.0769e-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(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4004e-05, 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(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2907e-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(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2441e-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(0.0007, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5479e-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(0.0003, 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(0.0004, 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(0.0005, 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(3.3548e-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(6.0403e-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(2.6887e-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(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-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(6.8303e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-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(8.0324e-05, 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(0.0007, 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(9.8018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2986e-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(7.4313e-05, 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(4.2540e-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(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-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(9.1824e-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(9.0251e-05, 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(0.0009, 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(5.7618e-05, device='cuda:0', grad_fn=<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(0.0003, 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(6.0080e-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(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5834e-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(0.0001, 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(3.9967e-05, device='cuda:0', grad_fn=<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(4.5447e-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(5.0109e-05, 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(4.0663e-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(0.0006, 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(8.2490e-05, device='cuda:0', grad_fn=<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(2.4435e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3755e-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(0.0002, 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(0.0002, 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(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(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-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(7.2269e-05, device='cuda:0', grad_fn=<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(0.0003, 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(4.5841e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<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(7.3935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2977e-05, 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(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8108e-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(9.8795e-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(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9199e-05, device='cuda:0', grad_fn=<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(0.0047, 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(8.4815e-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(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-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(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0173e-05, device='cuda:0', grad_fn=<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(5.3682e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0008, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-05, 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(6.1614e-05, device='cuda:0', grad_fn=<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(8.0233e-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(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5540e-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(0.0018, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-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(0.0004, 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(8.8375e-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(9.7064e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9548e-05, 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(0.0002, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-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(5.7164e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, 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(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(0.0001, 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(8.4877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-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(5.0079e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9817e-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(5.8799e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8233e-05, device='cuda:0', grad_fn=<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(7.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-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(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, 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(4.8687e-05, 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(8.9754e-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(0.0019, 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(0.0001, 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(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2230e-05, 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(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<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(9.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, 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(0.0005, 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(0.0008, 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(0.0003, 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(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(7.6388e-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(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(0.0007, 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(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-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(0.0057, 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(2.8795e-05, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_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(4.1784e-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(4.4932e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0002, 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(0.0002, 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(6.1039e-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(2.9945e-05, device='cuda:0', grad_fn=<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(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(6.1039e-05, 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(0.0006, 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(0.0003, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8500e-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(0.0001, 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(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-05, 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(9.7936e-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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5944e-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(0.0014, 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(4.6113e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-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(0.0004, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8507e-05, device='cuda:0', grad_fn=<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(6.6004e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0959e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8852e-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(7.5727e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-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(8.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0589e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0776e-05, device='cuda:0', grad_fn=<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(5.4590e-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(3.6455e-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(2.8643e-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(3.7000e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4999e-05, 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(6.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(4.0178e-05, device='cuda:0', grad_fn=<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(8.0201e-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(0.0008, 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(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<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(0.0012, 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(0.0004, 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(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-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(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1103e-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(8.8430e-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(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4488e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(8.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-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(7.2755e-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(3.4456e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4718e-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(0.0009, 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(0.0031, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-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(0.0024, 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
				 {'precision': {'support': 0.9413919413919414, 'agreement': 0.32786885245901637, 'direct_attack': 0.40625, 'undercutter_attack': 0.4470588235294118, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9145907473309609, 'agreement': 0.46511627906976744, 'direct_attack': 0.40625, 'undercutter_attack': 0.59375, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9277978339350181, 'agreement': 0.3846153846153846, 'direct_attack': 0.40625, 'undercutter_attack': 0.5100671140939598, 'partial': 0.044444444444444446}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 64, 'partial': 38}, 'micro_avg': {'precision': 0.6979591836734694, 'recall': 0.6979591836734694, 'f1': 0.6979591836734694, 'support': None}, 'macro_avg': {'precision': 0.45308535204750244, 'recall': 0.4812045631748825, 'f1': 0.45463495541776133, 'support': None}, 'weighted_avg': {'precision': 0.691162923104749, 'recall': 0.6979591836734694, 'f1': 0.6889445654389875, 'support': None}}
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(4.3873e-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(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-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(7.0030e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-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(0.0008, 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(0.0003, 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(5.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, 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(3.8362e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-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(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(0.0007, 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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5228e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0010, 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(0.0003, 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(7.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9139e-05, 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(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(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(6.7184e-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(0.0002, 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(5.0291e-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(6.9660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-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(7.3179e-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(0.0001, 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(4.7597e-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(3.8332e-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(9.0979e-05, device='cuda:0', grad_fn=<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(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-05, device='cuda:0', grad_fn=<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(6.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8458e-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(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-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(0.0006, 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(0.0009, 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(6.6428e-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(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6216e-05, 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(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(8.6533e-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(5.4954e-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(0.0001, 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(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-05, device='cuda:0', grad_fn=<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(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-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(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0291e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6382e-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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0001, 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(3.5456e-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(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7416e-05, 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(8.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1459e-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(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-05, device='cuda:0', grad_fn=<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(3.4608e-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(0.0008, 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(0.0001, 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(0.0007, 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(0.0003, 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(0.0001, 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(8.4424e-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(4.9988e-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(0.0004, 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(4.9050e-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(0.0002, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1196e-05, device='cuda:0', grad_fn=<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(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-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(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(4.0421e-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(9.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-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(0.0001, 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(4.8959e-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(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(9.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3340e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2079e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, 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(8.7316e-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(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4752e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4765e-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(0.0005, 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(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(0.0002, 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(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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-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(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4684e-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(0.0004, 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(7.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-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(9.3826e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5621e-05, device='cuda:0', grad_fn=<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(5.0987e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4058e-05, 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(9.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-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(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3257e-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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4576e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-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(9.8993e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-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(8.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5714e-05, device='cuda:0', grad_fn=<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(0.0002, 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(9.4096e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2173e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, 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(8.4381e-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(8.5713e-05, device='cuda:0', grad_fn=<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(9.0651e-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(7.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3591e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7876e-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(2.6887e-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(5.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6002e-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(4.6324e-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(0.0004, 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(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(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(7.5600e-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(0.0012, 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(0.0002, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-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(0.0007, 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(6.0343e-05, device='cuda:0', grad_fn=<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(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5670e-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(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2824e-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(6.5641e-05, 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(4.7142e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7357e-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(5.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-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(2.9340e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-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(5.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2959e-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(0.0001, 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(0.0008, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4953e-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(3.5062e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, 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(8.4940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1557e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3639e-05, device='cuda:0', grad_fn=<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(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, 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(0.0004, 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(8.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-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(0.0001, 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(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5450e-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(0.0007, 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(4.6416e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, 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(6.2461e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, 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(3.9664e-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(3.4608e-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(4.7808e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5167e-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(3.1338e-05, device='cuda:0', grad_fn=<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(3.6122e-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(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1459e-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(0.0001, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8625e-05, device='cuda:0', grad_fn=<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(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8561e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5847e-05, 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(8.4173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6964e-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(3.8120e-05, device='cuda:0', grad_fn=<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(2.6524e-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(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(7.7569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9021e-05, device='cuda:0', grad_fn=<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(8.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-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(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(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-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(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(4.2510e-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(0.0004, 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(0.0011, 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(0.0015, 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(0.0005, 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(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4002e-05, 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(5.5630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9637e-05, device='cuda:0', grad_fn=<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(9.1132e-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(4.7142e-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(3.6606e-05, device='cuda:0', grad_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(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(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-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(7.6212e-05, 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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-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(0.0003, 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(0.0005, 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(0.0005, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2353e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1079e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1762e-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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7072e-05, 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(8.9768e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-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(0.0003, 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(0.0002, 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(6.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-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(9.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1644e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2271e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9704e-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(0.0013, 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(0.0001, 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(0.0005, 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(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<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(7.8687e-05, 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(0.0007, 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(5.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(9.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271e-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(3.2882e-05, 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(3.8302e-05, device='cuda:0', grad_fn=<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(8.8998e-05, device='cuda:0', grad_fn=<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(7.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8668e-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(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(0.0018, 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(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4539e-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(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8527e-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(3.3911e-05, 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(0.0004, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-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(9.2494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-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(7.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-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(0.0001, 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(0.0004, 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(3.6394e-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(5.7836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-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(8.3045e-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(4.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-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(7.3269e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(9.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-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(0.0004, 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(3.1338e-05, 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(0.0020, 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(5.3016e-05, device='cuda:0', grad_fn=<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(3.0793e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9825e-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(7.1270e-05, 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(4.2450e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6538e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, 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(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-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(3.6576e-05, 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(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7169e-05, device='cuda:0', grad_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(6.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3913e-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(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-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(0.0006, 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(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1460e-05, device='cuda:0', grad_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(0.0017, 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(4.0300e-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(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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
				 {'precision': {'support': 0.9384057971014492, 'agreement': 0.34545454545454546, 'direct_attack': 0.38571428571428573, 'undercutter_attack': 0.4523809523809524, 'partial': 0.0}, 'recall': {'support': 0.9217081850533808, 'agreement': 0.4418604651162791, 'direct_attack': 0.421875, 'undercutter_attack': 0.5846153846153846, 'partial': 0.0}, 'f1': {'support': 0.9299820466786356, 'agreement': 0.3877551020408163, 'direct_attack': 0.4029850746268657, 'undercutter_attack': 0.5100671140939597, 'partial': 0.0}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 37}, 'micro_avg': {'precision': 0.7, 'recall': 0.7, 'f1': 0.7, 'support': None}, 'macro_avg': {'precision': 0.42439111613024655, 'recall': 0.4740118069570089, 'f1': 0.4461578674880554, 'support': None}, 'weighted_avg': {'precision': 0.6788511237357734, 'recall': 0.7, 'f1': 0.6876404728503644, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0148e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-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(2.6887e-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(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4248e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2735e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-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(2.2830e-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(0.0002, device='cuda:0', grad_fn=<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(8.5662e-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(3.2095e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7743e-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(0.0002, 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(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, 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(3.1096e-05, device='cuda:0', grad_fn=<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(3.7121e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4108e-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(6.0796e-05, 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(9.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0245e-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(8.9718e-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(0.0008, 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(7.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, 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(2.2800e-05, 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(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6315e-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(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0416e-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(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8871e-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(6.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-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(7.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5328e-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(4.5841e-05, 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(4.3872e-05, device='cuda:0', grad_fn=<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(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<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(6.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1945e-05, device='cuda:0', grad_fn=<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(2.8552e-05, 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(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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-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(2.6584e-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(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5691e-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(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5856e-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(9.6267e-05, device='cuda:0', grad_fn=<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(4.8474e-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(0.0013, 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(0.0003, 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(0.0013, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1734e-05, device='cuda:0', grad_fn=<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(4.6961e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-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(9.5997e-05, device='cuda:0', grad_fn=<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(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-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(4.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7435e-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(0.0031, device='cuda:0', grad_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(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7448e-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(9.3065e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0614e-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(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0029e-05, device='cuda:0', grad_fn=<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(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1956e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7354e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-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(3.7605e-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(0.0004, 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(6.4001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5367e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-05, device='cuda:0', grad_fn=<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(4.7475e-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(0.0001, 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(2.6796e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-05, device='cuda:0', grad_fn=<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(4.5931e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2097e-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(0.0006, 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(0.0007, 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(2.0347e-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(2.3829e-05, 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(2.3526e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0621e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9409e-05, 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(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-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(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4523e-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(2.9461e-05, device='cuda:0', grad_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-05, 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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-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(9.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3706e-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(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-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(3.0520e-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(0.0006, 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(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4822e-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(1.5291e-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(5.0442e-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(0.0001, 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(1.8833e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8919e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1219e-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(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0493e-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(8.0385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(6.4732e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<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(5.3288e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-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(9.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1864e-05, device='cuda:0', grad_fn=<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(1.8954e-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(8.7224e-05, 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(3.6334e-05, device='cuda:0', grad_fn=<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(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8565e-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(0.0001, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(9.6425e-05, device='cuda:0', grad_fn=<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(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5599e-05, device='cuda:0', grad_fn=<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(1.8228e-05, device='cuda:0', grad_fn=<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(8.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1071e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-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(3.2579e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2831e-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(2.9521e-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(7.6083e-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(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, 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(5.3470e-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(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(2.4586e-05, device='cuda:0', grad_fn=<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(0.0005, 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(2.4707e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(5.6549e-05, 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(0.0006, 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(7.4894e-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(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-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(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1034e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9287e-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(2.8764e-05, device='cuda:0', grad_fn=<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(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-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(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-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(6.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-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(1.7017e-05, device='cuda:0', grad_fn=<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(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-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(0.0004, 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(0.0007, 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(0.0001, 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(2.9703e-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(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5082e-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(0.0013, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-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(0.0004, 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(9.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7398e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, 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(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4448e-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(6.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5306e-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(1.7501e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6700e-05, 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(7.6376e-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(0.0005, 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(0.0004, 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(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(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(7.1453e-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(6.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6940e-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(5.3445e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9365e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9922e-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(0.0003, 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(7.6477e-05, device='cuda:0', grad_fn=<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(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-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(6.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, 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(2.6857e-05, 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(7.1157e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3245e-05, 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(2.6766e-05, device='cuda:0', grad_fn=<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(2.4768e-05, device='cuda:0', grad_fn=<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(8.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3211e-05, 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(0.0002, 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(7.2452e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8326e-05, device='cuda:0', grad_fn=<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(4.6052e-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(0.0013, 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(0.0002, 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(7.3430e-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(5.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, 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(0.0002, 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(8.0488e-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(0.0004, 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(4.4592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5125e-05, 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(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<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(0.0002, 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(8.4926e-05, device='cuda:0', grad_fn=<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(1.9348e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1431e-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(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7806e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3076e-05, 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(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7254e-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(4.9594e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9679e-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(6.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-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(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7140e-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(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6345e-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(0.0001, 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(0.0001, 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(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(7.4630e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0478e-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(0.0003, 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(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8657e-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(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, 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(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(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-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(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(2.4707e-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(4.2356e-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(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0528e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7887e-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(0.0002, 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(3.9997e-05, device='cuda:0', grad_fn=<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(8.1443e-05, device='cuda:0', grad_fn=<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(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-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(0.0005, 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(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<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(3.6273e-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(9.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4888e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0298e-05, 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(8.1623e-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(0.0001, 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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0167e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4756e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-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(6.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0196e-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(8.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4216e-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>)
			------------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
				 {'precision': {'support': 0.9384057971014492, 'agreement': 0.33962264150943394, 'direct_attack': 0.37681159420289856, 'undercutter_attack': 0.4470588235294118, 'partial': 0.14285714285714285}, 'recall': {'support': 0.925, 'agreement': 0.4186046511627907, 'direct_attack': 0.40625, 'undercutter_attack': 0.5846153846153846, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9316546762589928, 'agreement': 0.37499999999999994, 'direct_attack': 0.3909774436090226, 'undercutter_attack': 0.5066666666666666, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.6979591836734694, 'recall': 0.6979591836734694, 'f1': 0.6979591836734694, 'support': None}, 'macro_avg': {'precision': 0.44895119984006726, 'recall': 0.4721571650503719, 'f1': 0.4497486461958252, 'support': None}, 'weighted_avg': {'precision': 0.6856341505311839, 'recall': 0.6979591836734694, 'f1': 0.6870063019708523, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9557e-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(9.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-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(6.3567e-05, 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(0.0006, 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(6.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9202e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8485e-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(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3066e-05, device='cuda:0', grad_fn=<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(8.8502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, 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(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, 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(2.4253e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-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(0.0011, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2656e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-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(0.0004, 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(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<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(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2180e-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(9.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6774e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<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(5.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-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(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, 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(1.9893e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0571e-05, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2824e-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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7757e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, 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(7.1963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6260e-05, 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(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.8076e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4118e-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(3.8271e-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(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3884e-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(7.9504e-05, device='cuda:0', grad_fn=<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(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5668e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, 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(1.8228e-05, 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(5.7223e-05, device='cuda:0', grad_fn=<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(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-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(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-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(4.1814e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9685e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9239e-05, device='cuda:0', grad_fn=<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(6.6670e-05, device='cuda:0', grad_fn=<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(6.8266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(8.5290e-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(2.9854e-05, 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(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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4729e-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(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2571e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0409e-05, device='cuda:0', grad_fn=<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(4.8354e-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(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9636e-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(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-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(0.0003, 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(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4965e-05, 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(0.0002, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7354e-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(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, 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(3.9573e-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(7.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, 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(0.0002, 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(0.0004, 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(2.0196e-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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3456e-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(0.0001, 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(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-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(8.2759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6616e-05, 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(8.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6893e-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(0.0001, 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(2.5858e-05, 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(3.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1647e-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(0.0007, 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(0.0001, 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(5.4777e-05, 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(7.3536e-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(0.0001, 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(0.0004, 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(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5836e-05, device='cuda:0', grad_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(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(2.3163e-05, device='cuda:0', grad_fn=<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(8.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8533e-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(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(8.2981e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-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(6.0609e-05, device='cuda:0', grad_fn=<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(6.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<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(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8609e-05, 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(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2875e-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(7.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6748e-05, device='cuda:0', grad_fn=<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(0.0004, 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(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-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(2.8916e-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(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<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(3.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, 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(3.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7586e-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(2.1346e-05, device='cuda:0', grad_fn=<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(8.0452e-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(2.9673e-05, device='cuda:0', grad_fn=<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(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-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(0.0002, 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(3.1217e-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(2.8098e-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(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4849e-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(4.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5117e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4630e-05, 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(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(0.0001, 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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, 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(6.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-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(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7502e-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(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1194e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6877e-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(8.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9419e-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(9.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0006, 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(7.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4086e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-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(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4307e-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(6.8934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-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(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4117e-05, device='cuda:0', grad_fn=<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(4.8716e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1148e-05, 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(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(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0342e-05, device='cuda:0', grad_fn=<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(7.5206e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6893e-05, device='cuda:0', grad_fn=<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(2.3769e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-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(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-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(0.0004, 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(4.4084e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9350e-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(2.0620e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3237e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, 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(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6923e-05, 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(3.6061e-05, device='cuda:0', grad_fn=<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(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-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(6.2820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7892e-05, 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(0.0002, 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(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(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(4.7566e-05, 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(0.0009, 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(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9611e-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(3.0339e-05, device='cuda:0', grad_fn=<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(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7003e-05, device='cuda:0', grad_fn=<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(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5226e-05, 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(0.0002, device='cuda:0', grad_fn=<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(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-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(2.0347e-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(0.0003, 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(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8738e-05, device='cuda:0', grad_fn=<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(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8984e-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(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(1.9469e-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(7.4298e-05, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8129e-05, 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(6.1160e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<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(4.7693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-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(7.8025e-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(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3227e-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(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7859e-05, device='cuda:0', grad_fn=<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(4.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3422e-05, 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(2.3345e-05, device='cuda:0', grad_fn=<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(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2383e-05, device='cuda:0', grad_fn=<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(4.7990e-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(9.7407e-05, device='cuda:0', grad_fn=<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(8.6944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-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(6.0669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, 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(9.4785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0186e-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(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(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-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(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, 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(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8776e-05, device='cuda:0', grad_fn=<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(7.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-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(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8387e-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(0.0001, 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(0.0004, 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(4.9484e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-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(8.2039e-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(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(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-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(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4412e-05, 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(5.1108e-05, 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(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(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6529e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5850e-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(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0450e-05, device='cuda:0', grad_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>)
Loss: tensor(0.0005, 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(7.9351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-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(6.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-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(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-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(0.0005, 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(7.8031e-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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6147e-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(6.8199e-05, device='cuda:0', grad_fn=<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(0.0002, 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(9.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0489e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0846e-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(0.0004, 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(0.0004, 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(2.0166e-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(3.7660e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8003e-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(6.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-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(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-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
				 {'precision': {'support': 0.9384057971014492, 'agreement': 0.34615384615384615, 'direct_attack': 0.39705882352941174, 'undercutter_attack': 0.45977011494252873, 'partial': 0.14285714285714285}, 'recall': {'support': 0.925, 'agreement': 0.4186046511627907, 'direct_attack': 0.421875, 'undercutter_attack': 0.6153846153846154, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9316546762589928, 'agreement': 0.37894736842105264, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.5263157894736843, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7040816326530612, 'recall': 0.7040816326530612, 'f1': 0.7040816326530612, 'support': None}, 'macro_avg': {'precision': 0.4568491449168757, 'recall': 0.4814360112042181, 'f1': 0.45809063753781665, 'support': None}, 'weighted_avg': {'precision': 0.6905380248545699, 'recall': 0.7040816326530612, 'f1': 0.6923250603696322, 'support': None}}
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3797e-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(1.9075e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-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(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1071e-05, 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(7.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-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(0.0004, 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(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7688e-05, device='cuda:0', grad_fn=<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(8.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2292e-05, 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(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4697e-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(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(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5555e-05, device='cuda:0', grad_fn=<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(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0259e-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(0.0003, 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(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, 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(0.0004, 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(8.7026e-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(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-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(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(0.0002, 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(0.0012, 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(0.0003, 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(0.0001, 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(5.4567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-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(8.0327e-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(0.0004, 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(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-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(0.0003, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4302e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0051e-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(4.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, 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(6.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2523e-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(5.5655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1521e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4584e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2662e-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(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-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(6.6486e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3136e-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(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8191e-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(7.5116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6368e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(0.0002, 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(5.8995e-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(8.7940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9305e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7253e-05, 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(4.7045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9917e-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(0.0006, 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(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7282e-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(5.3756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-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(0.0003, 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(2.1710e-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(5.7343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-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(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(3.9391e-05, device='cuda:0', grad_fn=<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(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-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(5.5560e-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(2.5888e-05, 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(8.8831e-05, 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(7.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2352e-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(3.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4060e-05, device='cuda:0', grad_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(7.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0319e-05, device='cuda:0', grad_fn=<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(7.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-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(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, 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(0.0003, 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(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-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(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8724e-05, device='cuda:0', grad_fn=<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(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9705e-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(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8261e-05, device='cuda:0', grad_fn=<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(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-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(9.9034e-05, device='cuda:0', grad_fn=<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(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-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(1.4594e-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(0.0004, 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(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-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(5.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(5.7440e-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(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-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(0.0001, 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(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-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(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(6.2581e-05, 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(3.6818e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<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(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(2.3284e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8403e-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(6.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0002, 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(9.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2790e-05, 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(0.0003, 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(4.0770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3118e-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(4.9502e-05, device='cuda:0', grad_fn=<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(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-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(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-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(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0553e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1855e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-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(2.2648e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2481e-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(0.0006, 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(5.2410e-05, 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(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7142e-05, 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(4.7526e-05, device='cuda:0', grad_fn=<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(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-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(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2208e-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(2.1437e-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(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, 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(0.0001, 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(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4768e-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(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0649e-05, 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(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(5.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-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(0.0005, 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(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3742e-05, 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(7.1957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8825e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6125e-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(6.5679e-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(5.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3117e-05, 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(2.1982e-05, 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(0.0004, 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(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8536e-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(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9651e-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(2.1649e-05, 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(1.5957e-05, 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(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2784e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4771e-05, device='cuda:0', grad_fn=<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(3.4880e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4552e-05, device='cuda:0', grad_fn=<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(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-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(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(5.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4799e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3393e-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(6.4088e-05, device='cuda:0', grad_fn=<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(3.9053e-05, 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(3.5905e-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(4.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, 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(4.5113e-05, device='cuda:0', grad_fn=<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(0.0001, 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(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(3.5910e-05, device='cuda:0', grad_fn=<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(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8460e-05, 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(2.1256e-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(0.0003, 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(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_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(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5265e-05, device='cuda:0', grad_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(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6344e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0617e-05, 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(6.2095e-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(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(6.8606e-05, 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(2.8582e-05, device='cuda:0', grad_fn=<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(6.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4118e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7470e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1577e-05, device='cuda:0', grad_fn=<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(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-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(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-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(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(9.5612e-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(1.5624e-05, 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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9399e-05, device='cuda:0', grad_fn=<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(5.3408e-05, 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(8.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-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(0.0001, 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(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-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(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, 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(8.7954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9060e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-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(4.0011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-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(2.1801e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7057e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5195e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7766e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-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(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2767e-05, device='cuda:0', grad_fn=<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(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1198e-05, 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(0.0004, 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(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, 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(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(1.3928e-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(4.7899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0186e-05, 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(0.0002, device='cuda:0', grad_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(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0656e-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(1.4927e-05, device='cuda:0', grad_fn=<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(9.2681e-05, device='cuda:0', grad_fn=<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(6.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, 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(4.5416e-05, device='cuda:0', grad_fn=<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(5.9857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1309e-05, device='cuda:0', grad_fn=<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(9.4252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, 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(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(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(8.4321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4218e-05, device='cuda:0', grad_fn=<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(4.3478e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, 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(1.9893e-05, device='cuda:0', grad_fn=<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(2.4737e-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(0.0005, 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(0.0004, 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(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, 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(8.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3517e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, 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(2.1710e-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(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1309e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3844e-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(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-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(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<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(0.0003, 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(2.0953e-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(0.0004, 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(5.7347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6303e-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(9.2191e-05, 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
				 {'precision': {'support': 0.9386281588447654, 'agreement': 0.34615384615384615, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.4659090909090909, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9252669039145908, 'agreement': 0.4186046511627907, 'direct_attack': 0.421875, 'undercutter_attack': 0.640625, 'partial': 0.02631578947368421}, 'f1': {'support': 0.931899641577061, 'agreement': 0.37894736842105264, 'direct_attack': 0.4153846153846154, 'undercutter_attack': 0.5394736842105263, 'partial': 0.044444444444444446}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 64, 'partial': 38}, 'micro_avg': {'precision': 0.7081632653061225, 'recall': 0.7081632653061225, 'f1': 0.7081632653061225, 'support': None}, 'macro_avg': {'precision': 0.4605278295711509, 'recall': 0.4865374689102131, 'f1': 0.46202995080753995, 'support': None}, 'weighted_avg': {'precision': 0.6940157131603385, 'recall': 0.7081632653061225, 'f1': 0.6958333799759946, 'support': None}}
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<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(4.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.8197e-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(4.5871e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4671e-05, 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(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1186e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1646e-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(1.6260e-05, 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(0.0002, 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(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-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(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3318e-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(0.0003, 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(5.5127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8104e-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(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(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(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8943e-05, device='cuda:0', grad_fn=<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(4.2934e-05, 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(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6728e-05, device='cuda:0', grad_fn=<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(0.0001, 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(4.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7595e-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(9.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, 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(0.0009, 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(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3862e-05, 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(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2439e-05, 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(1.8197e-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(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2664e-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(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1746e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3593e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, 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(4.7358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0384e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5492e-05, device='cuda:0', grad_fn=<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(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2269e-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(0.0001, 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(2.2406e-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(3.1943e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(3.0430e-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(2.3890e-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(0.0007, 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(9.1771e-05, device='cuda:0', grad_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(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-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(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2101e-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(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1213e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2912e-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(5.1721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5168e-05, device='cuda:0', grad_fn=<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(7.8123e-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(2.8385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, 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(0.0003, 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(4.2789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5275e-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(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4138e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3897e-05, device='cuda:0', grad_fn=<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(0.0002, 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(9.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, 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(0.0003, 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(6.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5999e-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(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(8.3320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-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(0.0002, 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(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9615e-05, device='cuda:0', grad_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(3.5395e-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(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(1.9681e-05, 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(8.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2067e-05, device='cuda:0', grad_fn=<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(3.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, 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(3.5298e-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(1.4413e-05, device='cuda:0', grad_fn=<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(5.1430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8700e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-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(1.6229e-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(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0090e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9957e-05, device='cuda:0', grad_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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2574e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0510e-05, 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(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7262e-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(4.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-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(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3764e-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(0.0002, 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(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4230e-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(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-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(1.6653e-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(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(0.0001, 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(4.0299e-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(3.8331e-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(1.8803e-05, 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(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0283e-05, device='cuda:0', grad_fn=<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(2.1376e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4648e-05, device='cuda:0', grad_fn=<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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-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(0.0001, 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(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, 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(6.3610e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2178e-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(8.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8955e-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(7.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-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(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1457e-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(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4418e-05, device='cuda:0', grad_fn=<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(6.9700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6440e-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(7.3558e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2296e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1468e-05, 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(6.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2868e-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(9.5332e-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(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5275e-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(3.8847e-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(4.8047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6644e-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(9.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8585e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1078e-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(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5910e-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(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(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2409e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5177e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0101e-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(9.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9803e-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(8.2476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-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(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(4.5535e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<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(5.4234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0076e-05, 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(7.4547e-05, device='cuda:0', grad_fn=<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(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<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(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5057e-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(6.5352e-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(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-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(6.5916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-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(6.7543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8556e-05, 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(2.9824e-05, device='cuda:0', grad_fn=<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(8.2534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_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(4.1740e-05, 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(8.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(4.4689e-05, device='cuda:0', grad_fn=<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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6117e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<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(3.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5836e-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(3.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6617e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5122e-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(3.6727e-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(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(1.7289e-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(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, 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(7.8545e-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(8.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-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(3.5789e-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(7.1606e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-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(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, 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(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9920e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2945e-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(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, 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(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3416e-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(0.0003, 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(9.3511e-05, device='cuda:0', grad_fn=<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(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, 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(8.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4542e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-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(2.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8635e-05, device='cuda:0', grad_fn=<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(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-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(1.7380e-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(1.5351e-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(5.7011e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8317e-05, device='cuda:0', grad_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(9.8848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9317e-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(5.6195e-05, device='cuda:0', grad_fn=<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(3.0460e-05, 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(0.0001, 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(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-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(3.1580e-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(3.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5044e-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(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8878e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6827e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4935e-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(9.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-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(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3874e-05, 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(4.2134e-05, 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(8.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8761e-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(5.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6507e-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(3.0339e-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(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5067e-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(0.0003, 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(0.0005, 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(4.7748e-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(9.5373e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7119e-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(4.0155e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-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(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2182e-05, 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(5.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8981e-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(8.4408e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3213e-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(2.8189e-05, device='cuda:0', grad_fn=<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(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<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(4.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-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(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7318e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2814e-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(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1285e-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(1.9530e-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(1.1506e-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(0.0001, 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(2.3436e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1312e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9667e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-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(6.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8728e-05, 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(2.8310e-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(4.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, 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(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-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(1.9560e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9619e-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(4.5250e-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(7.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0484e-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(0.0003, 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>)
			------------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
				 {'precision': {'support': 0.9386281588447654, 'agreement': 0.3333333333333333, 'direct_attack': 0.3939393939393939, 'undercutter_attack': 0.4606741573033708, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9252669039145908, 'agreement': 0.40476190476190477, 'direct_attack': 0.40625, 'undercutter_attack': 0.6307692307692307, 'partial': 0.02631578947368421}, 'f1': {'support': 0.931899641577061, 'agreement': 0.3655913978494623, 'direct_attack': 0.4, 'undercutter_attack': 0.5324675324675324, 'partial': 0.044444444444444446}, 'support': {'support': 281, 'agreement': 42, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7040816326530612, 'recall': 0.7040816326530612, 'f1': 0.7040816326530612, 'support': None}, 'macro_avg': {'precision': 0.4538864372556012, 'recall': 0.47867276578388207, 'f1': 0.45488060326769997, 'support': None}, 'weighted_avg': {'precision': 0.690487807144471, 'recall': 0.7040816326530612, 'f1': 0.6920773805961431, 'support': None}}
Loss: tensor(3.5032e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0971e-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(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<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(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5600e-05, 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(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-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(2.6463e-05, device='cuda:0', grad_fn=<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(1.4080e-05, 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(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(1.5139e-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(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9043e-05, 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(8.1802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6021e-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(1.8167e-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(1.8985e-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(9.8390e-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(5.3010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4362e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-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(0.0001, 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(4.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0153e-05, device='cuda:0', grad_fn=<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(0.0002, 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(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8523e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-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(6.9042e-05, device='cuda:0', grad_fn=<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(8.3668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7695e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, 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(0.0001, 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(4.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2034e-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>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8036e-05, 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(6.0492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.6975e-05, device='cuda:0', grad_fn=<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(2.2890e-05, 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(3.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6018e-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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2918e-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(4.9486e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6080e-05, device='cuda:0', grad_fn=<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(4.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6458e-05, device='cuda:0', grad_fn=<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(7.7081e-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(0.0002, 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(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-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(2.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-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(9.4825e-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(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(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-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(0.0002, 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(3.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, 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(1.2990e-05, device='cuda:0', grad_fn=<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(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9887e-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(8.6862e-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(0.0001, 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(7.1271e-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(1.3262e-05, 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(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, 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(3.6061e-05, device='cuda:0', grad_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(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-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(9.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3142e-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(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(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9729e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1654e-05, 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(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7584e-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(0.0001, 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(3.0066e-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(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(1.5109e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3909e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1981e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0545e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0288e-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(3.5039e-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(6.2551e-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(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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2574e-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(3.3635e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3911e-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(5.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-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(0.0004, 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(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6118e-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(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5386e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-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(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4680e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3206e-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(5.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0733e-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(0.0002, 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(1.5109e-05, device='cuda:0', grad_fn=<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(0.0004, 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(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8932e-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(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(6.9667e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, 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(0.0001, 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(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-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(3.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-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(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0868e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8606e-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(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0256e-05, 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(0.0001, 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(1.4897e-05, device='cuda:0', grad_fn=<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(3.6651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8358e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2434e-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(4.9028e-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(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6262e-05, 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(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-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(3.8826e-05, 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(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9954e-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(6.8892e-05, 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(5.0155e-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(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-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(1.1294e-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(2.1649e-05, 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(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5526e-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(2.4435e-05, 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(1.4504e-05, 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(3.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<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(1.9651e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2051e-05, 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(0.0001, 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(6.5318e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, 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(6.3675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9271e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-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(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-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(3.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8263e-05, device='cuda:0', grad_fn=<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(5.1577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3887e-05, 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(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-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(1.5563e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2061e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4125e-05, device='cuda:0', grad_fn=<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(4.8785e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3489e-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(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5639e-05, device='cuda:0', grad_fn=<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(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-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(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4696e-05, device='cuda:0', grad_fn=<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(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1219e-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(4.1001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6848e-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(3.0797e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8191e-05, 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(0.0004, 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(6.3792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5076e-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(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3862e-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(0.0003, 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(1.1809e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3971e-05, device='cuda:0', grad_fn=<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(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, 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(6.2814e-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(2.2255e-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(3.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8161e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, 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(8.5770e-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(4.3206e-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(5.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5359e-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(3.4444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9730e-05, device='cuda:0', grad_fn=<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(9.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2127e-05, device='cuda:0', grad_fn=<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(4.8656e-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(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4022e-05, 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(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<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(8.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5566e-05, device='cuda:0', grad_fn=<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(2.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-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(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9030e-05, device='cuda:0', grad_fn=<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(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-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(5.1365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, 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(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(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8284e-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(0.0003, 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(9.7195e-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(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7553e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-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(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2998e-05, device='cuda:0', grad_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(3.7917e-05, 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(1.6653e-05, 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(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, 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(8.8933e-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(7.3667e-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(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0009e-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(4.6431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.3328e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-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(3.1005e-05, device='cuda:0', grad_fn=<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(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1285e-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(1.2354e-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(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-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(5.9642e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0621e-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(9.5604e-05, 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(7.2311e-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(8.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4259e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1874e-05, device='cuda:0', grad_fn=<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(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-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(2.6867e-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(0.0002, 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(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6804e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, 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(7.0398e-05, 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>)
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(3.7635e-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(6.3572e-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(0.0008, 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(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9426e-05, device='cuda:0', grad_fn=<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(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-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(7.7603e-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(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5221e-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(0.0003, 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(3.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-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(0.0001, 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(1.7319e-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(6.9618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-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(1.5927e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7987e-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(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, 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(3.7952e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_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(1.9045e-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(0.0001, 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(7.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1606e-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(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7081e-05, device='cuda: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 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
				 {'precision': {'support': 0.9384057971014492, 'agreement': 0.34615384615384615, 'direct_attack': 0.3939393939393939, 'undercutter_attack': 0.4606741573033708, 'partial': 0.14285714285714285}, 'recall': {'support': 0.925, 'agreement': 0.4186046511627907, 'direct_attack': 0.40625, 'undercutter_attack': 0.6307692307692307, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9316546762589928, 'agreement': 0.37894736842105264, 'direct_attack': 0.4, 'undercutter_attack': 0.5324675324675324, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7040816326530612, 'recall': 0.7040816326530612, 'f1': 0.7040816326530612, 'support': None}, 'macro_avg': {'precision': 0.4564060674710406, 'recall': 0.4813879342811411, 'f1': 0.45750280431840445, 'support': None}, 'weighted_avg': {'precision': 0.6902505131396589, 'recall': 0.7040816326530612, 'f1': 0.6919537238651057, 'support': None}}
Loss: tensor(1.0900e-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(6.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-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(6.7861e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7602e-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(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4656e-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(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(8.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7396e-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(8.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-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(7.1620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9195e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7662e-05, device='cuda:0', grad_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(7.3838e-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(7.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3959e-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(9.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0200e-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(6.7731e-05, device='cuda:0', grad_fn=<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(5.8779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-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(3.7241e-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(9.0632e-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(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(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1487e-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(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(1.3080e-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(6.1132e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9271e-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(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-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(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4372e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-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(9.7248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-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(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0548e-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(1.3565e-05, 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(4.8990e-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(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5650e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9480e-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(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(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<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(3.4396e-05, device='cuda:0', grad_fn=<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(5.7912e-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(1.5896e-05, device='cuda:0', grad_fn=<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(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, 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(4.7888e-05, device='cuda:0', grad_fn=<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(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6021e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3680e-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(1.0053e-05, 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(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4208e-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(5.7170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-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(7.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3528e-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(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, 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(3.2276e-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(7.8218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2058e-05, device='cuda:0', grad_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(2.3028e-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(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2850e-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(1.2929e-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(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(3.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6654e-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(6.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6530e-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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-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(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(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4696e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4446e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6687e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0240e-05, 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(0.0002, 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(0.0001, 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(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7695e-05, 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(1.8803e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4226e-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(1.4715e-05, 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(0.0001, 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(5.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6948e-05, 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(9.2101e-05, device='cuda:0', grad_fn=<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(2.8329e-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(9.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3362e-05, device='cuda:0', grad_fn=<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(9.2542e-05, device='cuda:0', grad_fn=<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(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7305e-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(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6891e-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(4.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7075e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.5810e-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(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, 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(7.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4830e-05, 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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0732e-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(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1969e-05, 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(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9992e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-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(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2983e-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(8.1042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8524e-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(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3673e-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(9.3511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4513e-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(3.3602e-05, 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(9.8215e-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(1.4170e-05, 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(5.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-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(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, 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(4.9848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2314e-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(2.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5340e-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(2.3570e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2042e-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(7.7437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5261e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1975e-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(9.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0986e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, 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(0.0001, 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(0.0003, 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(3.5411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8924e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-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(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-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(6.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0452e-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(7.4452e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6234e-05, device='cuda:0', grad_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(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-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(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5688e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3789e-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(6.4829e-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(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5552e-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(4.6998e-05, device='cuda:0', grad_fn=<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(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, 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(9.2334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3441e-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(0.0001, 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(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3714e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-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(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2063e-05, 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(8.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8110e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-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(1.5018e-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(3.0369e-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(9.2829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, 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(3.5923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6429e-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(8.2661e-06, 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(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9046e-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(4.5306e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9996e-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(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(6.6339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(4.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662e-05, 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(6.3366e-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(0.0001, 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(3.0577e-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(8.8325e-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(1.4231e-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(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3596e-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(0.0002, 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(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3341e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2346e-05, 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(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9562e-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(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, 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(4.9393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2267e-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(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8471e-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(1.8833e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8973e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2685e-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(1.6683e-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(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, 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(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2444e-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(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5480e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4201e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(9.5984e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6491e-05, device='cuda:0', grad_fn=<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(5.4270e-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(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5224e-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(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.1294e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3149e-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(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-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(4.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2809e-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(6.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3096e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0675e-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(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5181e-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(4.8465e-05, device='cuda:0', grad_fn=<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(3.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2616e-05, 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(2.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6786e-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(4.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8149e-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(7.8893e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4912e-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(7.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6420e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9767e-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(1.7501e-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(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3704e-05, device='cuda:0', grad_fn=<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(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8652e-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(1.2475e-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(4.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8940e-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(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2900e-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(0.0002, 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(4.7747e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8944e-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(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-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(5.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7743e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-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(2.5749e-05, 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(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2452e-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(1.2263e-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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-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(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5137e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_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(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9614e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3899e-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(6.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5778e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3389e-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(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9135e-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(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2987e-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
				 {'precision': {'support': 0.9386281588447654, 'agreement': 0.35294117647058826, 'direct_attack': 0.4, 'undercutter_attack': 0.4666666666666667, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9285714285714286, 'agreement': 0.4186046511627907, 'direct_attack': 0.40625, 'undercutter_attack': 0.6461538461538462, 'partial': 0.02631578947368421}, 'f1': {'support': 0.933572710951526, 'agreement': 0.3829787234042554, 'direct_attack': 0.40310077519379844, 'undercutter_attack': 0.5419354838709677, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7081632653061225, 'recall': 0.7081632653061225, 'f1': 0.7081632653061225, 'support': None}, 'macro_avg': {'precision': 0.4602186289678326, 'recall': 0.4851791430723499, 'f1': 0.4612064275729984, 'support': None}, 'weighted_avg': {'precision': 0.6925597139319885, 'recall': 0.7081632653061225, 'f1': 0.6950644676035003, 'support': None}}
Loss: tensor(5.9997e-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(3.6091e-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(8.8501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, 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(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1020e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8547e-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(3.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-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(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4847e-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(3.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, 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(8.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0315e-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(8.7007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1151e-05, device='cuda:0', grad_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(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(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-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(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3422e-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(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8858e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1955e-05, 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(5.1653e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3150e-05, device='cuda:0', grad_fn=<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(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9733e-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(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8067e-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(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4301e-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(2.1940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1914e-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(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0950e-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(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-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(4.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6525e-05, 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(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-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(8.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3735e-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(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9042e-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(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-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(9.4327e-05, 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(2.4533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6041e-05, 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(3.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(2.9491e-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(7.8725e-06, 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(9.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, 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(9.7801e-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(9.3900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3100e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6433e-05, 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(5.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3117e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, 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(1.0386e-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(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(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0276e-05, 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(6.3222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0285e-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(7.5058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2729e-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(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1467e-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(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3354e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8140e-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(9.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3294e-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(3.4807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4340e-05, device='cuda:0', grad_fn=<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(1.2747e-05, device='cuda:0', grad_fn=<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(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-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(6.1580e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1788e-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(6.4680e-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(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-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(9.0389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7440e-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(4.1631e-05, 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(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-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(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1635e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0235e-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(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1431e-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(4.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3981e-05, 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(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(8.2647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2354e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7824e-05, 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(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2968e-05, device='cuda:0', grad_fn=<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(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-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(6.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4665e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0963e-05, 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(1.1960e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9839e-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(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, 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(3.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, 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(1.6199e-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(0.0004, 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(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3017e-05, 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(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8798e-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(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-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(4.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1952e-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(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7072e-05, device='cuda:0', grad_fn=<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(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-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(1.8652e-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(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8950e-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(6.1053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, 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(6.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0037e-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(1.4170e-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(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3442e-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(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8861e-05, device='cuda:0', grad_fn=<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(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3506e-05, device='cuda:0', grad_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(9.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3689e-05, 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(8.5105e-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(8.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, 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(2.9907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3802e-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(6.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-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(4.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8569e-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(0.0002, 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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, 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(1.2414e-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(3.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8078e-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(2.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2680e-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(2.4876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8085e-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(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-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(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3273e-05, 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(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2434e-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(3.6895e-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(6.9944e-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(2.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3445e-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(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0183e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2747e-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(1.1294e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8658e-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(5.1049e-05, 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(7.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-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(8.9323e-06, 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(8.7547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0218e-05, 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(0.0002, 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(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7927e-05, device='cuda:0', grad_fn=<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(3.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2107e-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(2.7657e-05, device='cuda:0', grad_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(2.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6374e-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(8.5689e-06, 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(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2819e-05, 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(0.0001, 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(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4497e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-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(8.2896e-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(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5868e-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(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5526e-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(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-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(3.7798e-05, 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(8.0101e-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(2.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7692e-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(0.0002, 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(4.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2457e-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(2.7411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-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(6.6218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, 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(6.3199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4247e-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(6.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2765e-05, device='cuda:0', grad_fn=<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(1.0688e-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(2.4082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6733e-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(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0540e-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(7.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, 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(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8264e-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(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-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(8.0544e-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(0.0001, 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(7.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2606e-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(8.2304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7712e-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(8.4175e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4293e-05, 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(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5489e-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(2.0951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7605e-05, device='cuda:0', grad_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(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-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(0.0001, 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(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8947e-05, device='cuda:0', grad_fn=<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(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2788e-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(7.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7377e-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(7.9936e-06, 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(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(9.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9392e-05, device='cuda:0', grad_fn=<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(0.0006, 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(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5238e-05, device='cuda:0', grad_fn=<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(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0849e-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(3.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8209e-05, 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(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8781e-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(8.9625e-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(8.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-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(5.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-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(7.6762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, 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(9.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8494e-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(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6864e-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
				 {'precision': {'support': 0.9390681003584229, 'agreement': 0.36, 'direct_attack': 0.40625, 'undercutter_attack': 0.4666666666666667, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9323843416370107, 'agreement': 0.4186046511627907, 'direct_attack': 0.4126984126984127, 'undercutter_attack': 0.6461538461538462, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9357142857142856, 'agreement': 0.38709677419354843, 'direct_attack': 0.4094488188976378, 'undercutter_attack': 0.5419354838709677, 'partial': 0.044444444444444446}, 'support': {'support': 281, 'agreement': 43, 'direct_attack': 63, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7122448979591837, 'recall': 0.7122448979591837, 'f1': 0.7122448979591838, 'support': None}, 'macro_avg': {'precision': 0.4629683819764465, 'recall': 0.4872314082251489, 'f1': 0.4637279614241767, 'support': None}, 'weighted_avg': {'precision': 0.6953342672706563, 'recall': 0.7122448979591837, 'f1': 0.6985527479736526, 'support': None}}
Loss: tensor(0.0001, 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(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6786e-05, 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(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7806e-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(5.7247e-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(1.2172e-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(7.1327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9518e-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(6.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4331e-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(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6773e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9851e-05, 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(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2094e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5826e-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(1.1506e-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(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, 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(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3713e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5278e-05, 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(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5702e-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(6.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0526e-05, 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(8.3267e-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(2.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0410e-05, device='cuda:0', grad_fn=<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(2.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1858e-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(7.2249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9583e-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(5.0758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7375e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-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(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4371e-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(0.0002, 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(7.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-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(9.2351e-06, 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(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5465e-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(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-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(8.9881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-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(7.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0276e-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(2.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, 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(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9130e-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(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, 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(7.0266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5319e-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(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7784e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1001e-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(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7324e-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(9.4044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9469e-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(2.2335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-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(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-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(6.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-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(5.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8315e-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(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-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(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-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(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9611e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, 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(1.1597e-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(7.1516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0899e-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(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5848e-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(1.7347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7708e-05, 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(5.2773e-05, 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(9.6148e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286e-05, 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(7.7048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2384e-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(5.5509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9882e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3871e-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(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(6.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5749e-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(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6141e-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(8.2056e-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(3.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6279e-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(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0308e-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(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(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6289e-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(4.8215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9645e-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(8.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, 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(2.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6826e-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(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, 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(6.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4158e-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(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-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(6.3032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6139e-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(7.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-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(9.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4035e-05, 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(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7579e-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(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7562e-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(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2097e-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(2.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-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(8.5806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0300e-05, 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(4.4186e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3596e-05, 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(2.1612e-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(0.0002, 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(6.6203e-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(5.0216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1192e-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(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, 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(1.8767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2149e-05, 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(3.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, 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(5.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4110e-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(1.5835e-05, 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(7.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-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(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-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(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4036e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6662e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, 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(2.3937e-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(2.0468e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4692e-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(6.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, 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(9.7326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9177e-05, 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(8.5205e-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(9.2653e-06, 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(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7922e-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(1.6290e-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(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.9347e-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(7.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8143e-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(9.7763e-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(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1721e-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(5.9405e-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(8.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1913e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4259e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5371e-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(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8931e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-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(0.0001, 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(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-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(2.4444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1748e-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(7.6318e-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(4.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0713e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7877e-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(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7152e-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(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8779e-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(7.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2721e-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(1.3807e-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(7.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, 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(1.2717e-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(9.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8162e-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(2.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6914e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2675e-05, 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(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5630e-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(7.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2031e-05, device='cuda:0', grad_fn=<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(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3630e-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(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4159e-05, 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(4.2637e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1503e-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(1.2959e-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(0.0003, 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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8159e-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(0.0001, 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(5.9437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7185e-05, device='cuda:0', grad_fn=<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(5.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6168e-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(1.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5454e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4244e-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(8.6270e-05, 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(5.6544e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2567e-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(2.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6233e-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(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-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(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-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(0.0002, 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(6.3852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, 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(5.3109e-05, device='cuda:0', grad_fn=<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(6.8128e-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(8.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3639e-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(3.4120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, 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(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8945e-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(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(9.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5966e-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(3.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-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(4.1420e-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(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-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(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(3.2573e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0556e-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(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2962e-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(8.3788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6140e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-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(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2832e-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(9.8127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0524e-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
				 {'precision': {'support': 0.9388489208633094, 'agreement': 0.3673469387755102, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.4666666666666667, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9321428571428572, 'agreement': 0.4186046511627907, 'direct_attack': 0.421875, 'undercutter_attack': 0.6461538461538462, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9354838709677419, 'agreement': 0.391304347826087, 'direct_attack': 0.4153846153846154, 'undercutter_attack': 0.5419354838709677, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7122448979591837, 'recall': 0.7122448979591837, 'f1': 0.7122448979591838, 'support': None}, 'macro_avg': {'precision': 0.46496211565070766, 'recall': 0.4890184287866356, 'f1': 0.46571055249877125, 'support': None}, 'weighted_avg': {'precision': 0.6951374268424418, 'recall': 0.7122448979591837, 'f1': 0.6984915950053197, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2723e-05, 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(5.6921e-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(5.6781e-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(6.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8045e-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(2.3355e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-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(5.7533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1876e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, 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(8.7966e-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(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, 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(3.3783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-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(7.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1071e-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(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9461e-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(7.9990e-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(6.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2049e-05, 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(1.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0467e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-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(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0594e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3302e-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(5.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9973e-05, 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(6.1503e-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(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1124e-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(7.5539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5951e-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(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7370e-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(7.1156e-06, 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(1.5532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5229e-05, 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(4.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9489e-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(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7039e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-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(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3242e-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(7.4947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1889e-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(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4221e-05, 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(0.0001, 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(2.8643e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1568e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, 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(1.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9526e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7388e-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(7.9960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4457e-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(2.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0012e-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(6.2578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, 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(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7029e-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(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5196e-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(7.9634e-06, 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(1.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3806e-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(5.9804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9164e-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(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-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(1.0870e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, 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(2.1764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-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(7.9174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6704e-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(2.8491e-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(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2547e-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(7.9331e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6407e-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(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2996e-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(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6017e-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(1.0840e-05, 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(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-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(9.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1698e-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(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4522e-05, device='cuda:0', grad_fn=<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(9.0865e-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(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-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(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(7.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0336e-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(0.0001, 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(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1721e-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(6.6522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7343e-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(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6864e-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(5.9418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1541e-05, 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(4.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0501e-05, 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(9.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9841e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-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(9.8744e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5857e-05, 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(2.3216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2426e-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(9.6287e-06, 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(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6612e-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(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8397e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5883e-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(6.5483e-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(9.6401e-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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7887e-05, device='cuda:0', grad_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(7.6508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2143e-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(7.3174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-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(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, 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(3.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6050e-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(8.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-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(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8391e-05, 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(8.6200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-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(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8935e-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(1.8561e-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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8920e-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(9.7549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1744e-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(1.4297e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6894e-05, 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(3.7564e-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(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, 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(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, 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(8.1666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, 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(4.4263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5271e-05, device='cuda:0', grad_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(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9215e-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(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9406e-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(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-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(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1960e-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(2.4737e-05, device='cuda:0', grad_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(2.7774e-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(9.0519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2623e-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(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0701e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6725e-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(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6793e-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(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8100e-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(6.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9459e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1907e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8767e-05, device='cuda:0', grad_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(8.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-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(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-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(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, 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(4.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6337e-05, 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(3.0692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-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
				 {'precision': {'support': 0.9388489208633094, 'agreement': 0.3673469387755102, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.4666666666666667, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9321428571428572, 'agreement': 0.4186046511627907, 'direct_attack': 0.421875, 'undercutter_attack': 0.6461538461538462, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9354838709677419, 'agreement': 0.391304347826087, 'direct_attack': 0.4153846153846154, 'undercutter_attack': 0.5419354838709677, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7122448979591837, 'recall': 0.7122448979591837, 'f1': 0.7122448979591838, 'support': None}, 'macro_avg': {'precision': 0.46496211565070766, 'recall': 0.4890184287866356, 'f1': 0.46571055249877125, 'support': None}, 'weighted_avg': {'precision': 0.6951374268424418, 'recall': 0.7122448979591837, 'f1': 0.6984915950053197, 'support': None}}
Loss: tensor(9.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-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(5.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3004e-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(7.8188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(2.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4777e-05, device='cuda:0', grad_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(3.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2244e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7430e-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(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7936e-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(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2110e-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(1.6621e-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(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-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(6.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-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(7.5726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6621e-05, 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(4.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-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(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, 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(2.4587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5946e-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(4.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8049e-05, device='cuda:0', grad_fn=<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(4.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8559e-05, 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(9.4506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2693e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-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(1.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3971e-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(6.5028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1970e-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(4.2054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1252e-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(1.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9354e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7017e-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(4.6370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2769e-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(2.2639e-05, 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(1.2596e-05, 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(8.2296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9912e-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(8.6372e-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(6.3152e-05, 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(2.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<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(8.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9703e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6319e-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(1.7347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7667e-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(5.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1122e-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(5.3514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9081e-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(4.4195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, 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(7.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4053e-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(5.5044e-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(3.4675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-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(3.9004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, 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(3.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7933e-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(8.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8981e-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(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, 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(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3764e-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(3.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4062e-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(7.0173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7496e-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(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-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(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0546e-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(4.6280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2843e-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(4.8052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6461e-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(4.5838e-05, device='cuda:0', grad_fn=<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(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8628e-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(1.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9973e-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(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0661e-05, 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(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2668e-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(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5161e-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(1.3965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-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(2.6675e-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(2.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9550e-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(3.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8783e-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(9.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8815e-05, 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(1.0325e-05, 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(5.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-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(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-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(7.9541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-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(1.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7573e-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(6.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4266e-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(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-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(4.4137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8978e-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(1.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2115e-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(5.0225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6856e-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(2.9139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5681e-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(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-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(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3145e-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(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(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2573e-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(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4900e-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(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7937e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, 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(4.8637e-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(4.2535e-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(7.7119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4812e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, 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(5.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7643e-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(1.7744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8318e-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(9.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7331e-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(8.3570e-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(2.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2693e-05, 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(4.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6526e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0474e-05, 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(2.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4472e-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(6.7522e-06, 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(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, 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(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7557e-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(4.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0683e-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(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-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(6.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1581e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0809e-05, 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(1.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5893e-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(5.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6258e-05, 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
				 {'precision': {'support': 0.9386281588447654, 'agreement': 0.3673469387755102, 'direct_attack': 0.4153846153846154, 'undercutter_attack': 0.4725274725274725, 'partial': 0.125}, 'recall': {'support': 0.9285714285714286, 'agreement': 0.4186046511627907, 'direct_attack': 0.421875, 'undercutter_attack': 0.6615384615384615, 'partial': 0.02631578947368421}, 'f1': {'support': 0.933572710951526, 'agreement': 0.391304347826087, 'direct_attack': 0.4186046511627907, 'undercutter_attack': 0.5512820512820512, 'partial': 0.043478260869565216}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7122448979591837, 'recall': 0.7122448979591837, 'f1': 0.7122448979591838, 'support': None}, 'macro_avg': {'precision': 0.4637774371064727, 'recall': 0.49138106614927296, 'f1': 0.46764840441840405, 'support': None}, 'weighted_avg': {'precision': 0.6952259264138415, 'recall': 0.7122448979591837, 'f1': 0.6989850019260091, 'support': None}}
Loss: tensor(6.4938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-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(3.8903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8496e-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(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2610e-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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1050e-05, device='cuda:0', grad_fn=<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(7.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, 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(5.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1916e-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(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, 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(8.2758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8481e-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(7.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1333e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4506e-05, 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(1.7558e-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(2.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-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(3.5128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7546e-05, 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(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0211e-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(7.0605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, 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(3.2594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3112e-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(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, 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(6.0064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2816e-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(0.0001, 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(3.1614e-05, 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(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-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(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4263e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0526e-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(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6432e-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(6.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6275e-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(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, 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(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0650e-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(1.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4866e-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(4.3870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1377e-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(8.8602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9155e-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(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8884e-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(8.5037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-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(5.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0441e-05, device='cuda:0', grad_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(5.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6350e-05, 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(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9916e-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(5.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, 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(6.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6531e-05, 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(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, 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(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-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(1.6443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4074e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-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(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7638e-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(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, 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(8.3879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3113e-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(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8393e-05, 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(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3344e-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(5.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3044e-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(2.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9594e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-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(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6739e-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(6.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7037e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3230e-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(6.2896e-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(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7229e-05, 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(1.9731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5608e-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(9.3788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0877e-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(6.6680e-05, 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(5.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0471e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-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(4.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-05, 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(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-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(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6807e-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(4.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6526e-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(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9114e-05, 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(4.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, 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(7.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7434e-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(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-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(4.0861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(6.4785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2699e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0666e-05, 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(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5209e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7591e-05, device='cuda:0', grad_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(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6563e-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(3.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, 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(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7882e-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
				 {'precision': {'support': 0.9390681003584229, 'agreement': 0.375, 'direct_attack': 0.421875, 'undercutter_attack': 0.4673913043478261, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9323843416370107, 'agreement': 0.42857142857142855, 'direct_attack': 0.421875, 'undercutter_attack': 0.6615384615384615, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9357142857142856, 'agreement': 0.39999999999999997, 'direct_attack': 0.421875, 'undercutter_attack': 0.5477707006369427, 'partial': 0.044444444444444446}, 'support': {'support': 281, 'agreement': 42, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7163265306122449, 'recall': 0.7163265306122449, 'f1': 0.7163265306122448, 'support': None}, 'macro_avg': {'precision': 0.4692383095126783, 'recall': 0.494137004244117, 'f1': 0.46996088615913456, 'support': None}, 'weighted_avg': {'precision': 0.6988513110446877, 'recall': 0.7163265306122449, 'f1': 0.7021014259510294, 'support': None}}
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1575e-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(3.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7895e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, 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(9.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7422e-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(2.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6198e-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(4.6582e-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(4.5452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0958e-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(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3166e-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(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, 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(8.0431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5303e-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(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8914e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1300e-05, 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(3.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9742e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4666e-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(3.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5475e-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(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6903e-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(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5485e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3677e-05, 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(1.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1313e-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(3.6268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9850e-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(6.3183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6708e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6739e-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(4.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0655e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0329e-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(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9453e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2713e-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(6.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8565e-05, device='cuda:0', grad_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(2.3450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-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(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3393e-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(1.2364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1636e-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(1.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-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(4.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7794e-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(5.4311e-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(3.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7421e-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(4.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1522e-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(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0118e-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(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2330e-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(4.0062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1662e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, 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(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-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(1.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6535e-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(1.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7213e-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(4.3489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6612e-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(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-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(4.5293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1665e-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(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-05, 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(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, 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(6.3282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7373e-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(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-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(3.0253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2805e-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(1.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2484e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3614e-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(5.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, 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(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0997e-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(3.7633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-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(8.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7463e-05, 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(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0420e-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(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, 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(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0029e-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(1.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3111e-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(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6416e-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(1.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2810e-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(7.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-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(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7952e-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(1.6327e-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(4.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6514e-05, 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(1.7554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4632e-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(7.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9354e-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(1.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9288e-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(1.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, 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(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8500e-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(8.0410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5212e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2299e-05, 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(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-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(3.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7878e-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
				 {'precision': {'support': 0.9388489208633094, 'agreement': 0.375, 'direct_attack': 0.4090909090909091, 'undercutter_attack': 0.4725274725274725, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9321428571428572, 'agreement': 0.4186046511627907, 'direct_attack': 0.421875, 'undercutter_attack': 0.6615384615384615, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9354838709677419, 'agreement': 0.3956043956043956, 'direct_attack': 0.4153846153846154, 'undercutter_attack': 0.5512820512820512, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7142857142857143, 'recall': 0.7142857142857143, 'f1': 0.7142857142857143, 'support': None}, 'macro_avg': {'precision': 0.46766488906776676, 'recall': 0.4920953518635587, 'f1': 0.46843987553664973, 'support': None}, 'weighted_avg': {'precision': 0.6965864758497999, 'recall': 0.7142857142857143, 'f1': 0.7001087969159069, 'support': None}}
Loss: tensor(1.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, 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(9.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5494e-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(1.3995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1643e-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(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8591e-05, 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(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5108e-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(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7965e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7494e-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(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1354e-05, 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(3.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8205e-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(1.7753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4182e-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(1.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0088e-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(6.7559e-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(3.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8257e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5856e-05, 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(8.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2813e-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(4.0086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3234e-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(7.8265e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, 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(4.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9922e-05, 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(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2460e-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(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6382e-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(1.5086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5517e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4309e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8862e-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(2.3019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1121e-05, 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(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, 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(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3983e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4301e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7746e-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(1.2242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, 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(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-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(5.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3101e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7325e-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(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1269e-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(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9774e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7592e-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(2.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0666e-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(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8816e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3424e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4675e-05, 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(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4333e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4607e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7776e-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(7.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9427e-05, 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(4.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9742e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, 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(1.7865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-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(3.3913e-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(1.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, 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(4.5009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9430e-05, device='cuda:0', grad_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(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9331e-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(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1790e-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(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, 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(1.1456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9445e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6734e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3731e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3683e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0221e-05, 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(3.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6126e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, 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(3.5172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1289e-05, 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(5.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-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
				 {'precision': {'support': 0.9388489208633094, 'agreement': 0.3673469387755102, 'direct_attack': 0.4, 'undercutter_attack': 0.4725274725274725, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9321428571428572, 'agreement': 0.4186046511627907, 'direct_attack': 0.40625, 'undercutter_attack': 0.6615384615384615, 'partial': 0.02631578947368421}, 'f1': {'support': 0.9354838709677419, 'agreement': 0.391304347826087, 'direct_attack': 0.40310077519379844, 'undercutter_attack': 0.5512820512820512, 'partial': 0.044444444444444446}, 'support': {'support': 280, 'agreement': 43, 'direct_attack': 64, 'undercutter_attack': 65, 'partial': 38}, 'micro_avg': {'precision': 0.7122448979591837, 'recall': 0.7122448979591837, 'f1': 0.7122448979591838, 'support': None}, 'macro_avg': {'precision': 0.46431609500468707, 'recall': 0.4889703518635587, 'f1': 0.4651230979428246, 'support': None}, 'weighted_avg': {'precision': 0.6947274966365933, 'recall': 0.7122448979591837, 'f1': 0.6981270258410506, '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
				 {'precision': {'support': 0.6302521008403361, 'agreement': 0.0, 'direct_attack': 0.12962962962962962, 'undercutter_attack': 0.12700729927007298, 'partial': 0.06794055201698514}, 'recall': {'support': 0.08971291866028708, 'agreement': 0.0, 'direct_attack': 0.12727272727272726, 'undercutter_attack': 0.5209580838323353, 'partial': 0.32989690721649484}, 'f1': {'support': 0.15706806282722513, 'agreement': 0.0, 'direct_attack': 0.12844036697247707, 'undercutter_attack': 0.20422535211267603, 'partial': 0.11267605633802817}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.15007215007215008, 'recall': 0.15007215007215008, 'f1': 0.15007215007215008, 'support': None}, 'macro_avg': {'precision': 0.19096591635140475, 'recall': 0.2135681273963689, 'f1': 0.12048196765008128, 'support': None}, 'weighted_avg': {'precision': 0.41049817322188314, 'recall': 0.15007215007215008, 'f1': 0.1374260838081806, 'support': None}}
Loss: tensor(2.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3876, 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
				 {'precision': {'support': 0.8825466520307355, 'agreement': 0.8977272727272727, 'direct_attack': 0.2037914691943128, 'undercutter_attack': 0.5194805194805194, 'partial': 0.18181818181818182}, 'recall': {'support': 0.9617224880382775, 'agreement': 0.44886363636363635, 'direct_attack': 0.39090909090909093, 'undercutter_attack': 0.47904191616766467, 'partial': 0.041237113402061855}, 'f1': {'support': 0.9204350314825415, 'agreement': 0.5984848484848484, 'direct_attack': 0.2679127725856698, 'undercutter_attack': 0.4984423676012461, 'partial': 0.06722689075630252}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7287157287157288, 'recall': 0.7287157287157288, 'f1': 0.7287157287157288, 'support': None}, 'macro_avg': {'precision': 0.5370728190502044, 'recall': 0.46435484897614626, 'f1': 0.4705003821821217, 'support': None}, 'weighted_avg': {'precision': 0.7378179459586433, 'recall': 0.7287157287157288, 'f1': 0.717206571738767, 'support': None}}
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3751, 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
				 {'precision': {'support': 0.9201474201474201, 'agreement': 0.4636363636363636, 'direct_attack': 0.2422907488986784, 'undercutter_attack': 0.58, 'partial': 0.32}, 'recall': {'support': 0.895933014354067, 'agreement': 0.5795454545454546, 'direct_attack': 0.5, 'undercutter_attack': 0.3473053892215569, 'partial': 0.08247422680412371}, 'f1': {'support': 0.907878787878788, 'agreement': 0.5151515151515151, 'direct_attack': 0.3264094955489614, 'undercutter_attack': 0.4344569288389513, 'partial': 0.13114754098360656}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7012987012987013, 'recall': 0.7012987012987013, 'f1': 0.7012987012987013, 'support': None}, 'macro_avg': {'precision': 0.5052149065364924, 'recall': 0.48105161698504045, 'f1': 0.4630088536803645, 'support': None}, 'weighted_avg': {'precision': 0.7253933806797243, 'recall': 0.7012987012987013, 'f1': 0.70045742888545, 'support': None}}
Loss: tensor(2.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, 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(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0146, 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(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, 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(0.2704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, 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
				 {'precision': {'support': 0.8841394825646794, 'agreement': 0.5314685314685315, 'direct_attack': 0.20408163265306123, 'undercutter_attack': 0.4727272727272727, 'partial': 0.15897435897435896}, 'recall': {'support': 0.9401913875598086, 'agreement': 0.4318181818181818, 'direct_attack': 0.09090909090909091, 'undercutter_attack': 0.31137724550898205, 'partial': 0.31958762886597936}, 'f1': {'support': 0.9113043478260869, 'agreement': 0.47648902821316613, 'direct_attack': 0.12578616352201258, 'undercutter_attack': 0.37545126353790614, 'partial': 0.21232876712328766}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.689033189033189, 'recall': 0.689033189033189, 'f1': 0.689033189033189, 'support': None}, 'macro_avg': {'precision': 0.45027825567758073, 'recall': 0.41877670693240854, 'f1': 0.4202719140444919, 'support': None}, 'weighted_avg': {'precision': 0.6850606175471412, 'recall': 0.689033189033189, 'f1': 0.6802635159865342, 'support': None}}
Loss: tensor(0.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, 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(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, 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(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, 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(1.8277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2889, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(2.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7253, 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(0.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7986, 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(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, 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(0.0075, 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(1.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9520, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7254, 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(1.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3567, 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(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5251, 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(1.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1520, 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(0.8444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7667, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0260, 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
				 {'precision': {'support': 0.8820286659316428, 'agreement': 0.6229508196721312, 'direct_attack': 0.2391304347826087, 'undercutter_attack': 0.45528455284552843, 'partial': 0.24615384615384617}, 'recall': {'support': 0.9569377990430622, 'agreement': 0.4318181818181818, 'direct_attack': 0.1, 'undercutter_attack': 0.6706586826347305, 'partial': 0.16494845360824742}, 'f1': {'support': 0.9179575444635685, 'agreement': 0.5100671140939597, 'direct_attack': 0.14102564102564102, 'undercutter_attack': 0.5423728813559321, 'partial': 0.19753086419753088}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7323232323232324, 'recall': 0.7323232323232324, 'f1': 0.7323232323232324, 'support': None}, 'macro_avg': {'precision': 0.4891096638771514, 'recall': 0.46487262342084434, 'f1': 0.46179080902732644, 'support': None}, 'weighted_avg': {'precision': 0.7021854979865524, 'recall': 0.7323232323232324, 'f1': 0.7088267711244601, 'support': None}}
Loss: tensor(0.0389, 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(0.9172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6042, 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(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9166, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3228, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, 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(0.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4056, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, 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(0.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5620, 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(1.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6283, 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(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, 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(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3042, 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(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7581, 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(2.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, 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(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, 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(0.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9718, 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(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7509, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6866, 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(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, 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(2.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8526, 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(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1738, 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(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9094, 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(0.0578, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(1.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, 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(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, 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(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, 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(0.0558, 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(1.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, 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(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4943, 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(1.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, 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(4.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, 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(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, 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(1.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8795, 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(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6761, 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(1.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4202, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3705, 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(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6521, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.4700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5991, 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(1.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3572, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8467, 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(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7709, 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(0.0025, 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(0.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1014, 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(0.0069, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(1.3163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(3.0380, 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(0.3452, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2424, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8596, 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(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2901, 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
				 {'precision': {'support': 0.8885209713024282, 'agreement': 0.7168141592920354, 'direct_attack': 0.2217573221757322, 'undercutter_attack': 0.46534653465346537, 'partial': 0.2222222222222222}, 'recall': {'support': 0.9629186602870813, 'agreement': 0.4602272727272727, 'direct_attack': 0.4818181818181818, 'undercutter_attack': 0.281437125748503, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9242250287026406, 'agreement': 0.5605536332179931, 'direct_attack': 0.3037249283667622, 'undercutter_attack': 0.3507462686567164, 'partial': 0.0967741935483871}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7157287157287158, 'recall': 0.7157287157287158, 'f1': 0.7157287157287159, 'support': None}, 'macro_avg': {'precision': 0.5029322419291767, 'recall': 0.4496513821368264, 'f1': 0.44720481049849986, 'support': None}, 'weighted_avg': {'precision': 0.7161793335687179, 'recall': 0.7157287157287158, 'f1': 0.7017900643592953, 'support': None}}
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9248, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5229, 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(1.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, 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(1.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8763, 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(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2227, 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(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, 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(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, 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(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3270, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, 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(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, 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(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9176, 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(2.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2153, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, 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(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, 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(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2977, 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(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0728, 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(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2382, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4753, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3824, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9123, 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(0.0785, 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(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0482, 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(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3949, 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(0.7720, 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(0.8091, 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(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, 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(0.0515, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6085, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3139, 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(0.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, 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(0.0478, 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(1.7603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, 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(1.4253, 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(0.0144, 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(0.0014, 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(0.7301, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, 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(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, 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(1.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0395, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, 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(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, 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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, 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(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2292, 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(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7735, 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(0.4367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, 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(0.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, 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(0.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, 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(0.2318, 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(0.4893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6567, 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(0.5270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, 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(0.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, 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(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4845, 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(0.0006, 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(0.2300, 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(0.0381, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1873, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, 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(0.1769, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, 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(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, 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(0.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5031, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, 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(0.3798, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, 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(0.0006, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, 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(1.6681, 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(0.0087, 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(0.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, 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(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275, 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(0.0556, 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(0.0074, 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(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1197, 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(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, 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(0.0332, 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(0.0029, 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(0.0033, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4431, 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(0.5043, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, 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(0.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0007, 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(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, 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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, 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(0.1115, 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(0.0024, 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(2.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, 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(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, 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(0.0356, 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(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, 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(0.0007, 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(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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
				 {'precision': {'support': 0.9138134592680047, 'agreement': 0.5240641711229946, 'direct_attack': 0.22304832713754646, 'undercutter_attack': 0.47619047619047616, 'partial': 0.17073170731707318}, 'recall': {'support': 0.9258373205741627, 'agreement': 0.5568181818181818, 'direct_attack': 0.5454545454545454, 'undercutter_attack': 0.11976047904191617, 'partial': 0.07216494845360824}, 'f1': {'support': 0.9197860962566845, 'agreement': 0.5399449035812672, 'direct_attack': 0.31662269129287596, 'undercutter_attack': 0.19138755980861247, 'partial': 0.10144927536231883}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.6919191919191919, 'recall': 0.6919191919191919, 'f1': 0.6919191919191919, 'support': None}, 'macro_avg': {'precision': 0.46156962820721903, 'recall': 0.4440070950684829, 'f1': 0.4138381052603518, 'support': None}, 'weighted_avg': {'precision': 0.7047643919079326, 'recall': 0.6919191919191919, 'f1': 0.6786452220355633, 'support': None}}
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, 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(0.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, 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(0.0308, 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(0.0125, 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(0.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, 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(0.0080, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.9060, 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(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6500, 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(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0002, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4375, 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(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9620, 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(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, 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(0.0007, 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(0.0131, 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(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1606, 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(0.1140, 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(0.0012, 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(0.1906, 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(0.0162, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9618, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, 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(0.0134, 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(0.0050, 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(0.0002, 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(0.0015, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, 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(0.0007, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2692, 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(0.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825, 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(0.0026, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, 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(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5759, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, 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(0.1864, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<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(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, 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(0.0004, 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(0.0003, 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(0.1000, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6045, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, 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(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2549, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4389, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, 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(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, 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(0.0107, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, 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(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, 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(0.1919, 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(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, 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(5.7801, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0127, 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(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4497, 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(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0092, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, 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(0.2413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, 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(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, 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(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.1284, 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(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, 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(0.3611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, 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(0.0906, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, 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(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, 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(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, 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(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, 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(0.0135, 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(0.0015, 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(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4374, 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(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2242, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6583, 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(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4886, 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(0.0185, 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(0.0007, 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(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, 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(0.9320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, 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(0.1168, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, 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(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, 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(0.0130, 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(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, 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(0.0140, 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(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, 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(0.7331, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7622, 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(0.0008, 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(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, 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(0.0084, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.3772, 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(4.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4473, 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(0.0512, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5228, 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(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7266, 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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, 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(0.0009, 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>)
			------------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
				 {'precision': {'support': 0.8427543679342241, 'agreement': 0.7727272727272727, 'direct_attack': 0.21481481481481482, 'undercutter_attack': 0.4748603351955307, 'partial': 0.2909090909090909}, 'recall': {'support': 0.9808612440191388, 'agreement': 0.19318181818181818, 'direct_attack': 0.2636363636363636, 'undercutter_attack': 0.5089820359281437, 'partial': 0.16494845360824742}, 'f1': {'support': 0.9065782200110559, 'agreement': 0.3090909090909091, 'direct_attack': 0.236734693877551, 'undercutter_attack': 0.49132947976878616, 'partial': 0.21052631578947367}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.70995670995671, 'recall': 0.70995670995671, 'f1': 0.70995670995671, 'support': None}, 'macro_avg': {'precision': 0.5192131763161866, 'recall': 0.42232198307474234, 'f1': 0.43085192370755515, 'support': None}, 'weighted_avg': {'precision': 0.7010765793784102, 'recall': 0.70995670995671, 'f1': 0.6787974632097687, 'support': None}}
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, 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(0.0078, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, 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(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, 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(0.0021, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2644, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.3300, 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(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, 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(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, 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(0.0005, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1716, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, 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(0.0260, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, 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(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, 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(0.1235, 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(0.0574, 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(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, 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(0.0113, 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(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, 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(0.0195, 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(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, 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(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(2.1370, 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(0.0016, 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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0146, 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(0.0005, 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(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, 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(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244, 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(0.0085, 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(0.0271, 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(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7259, device='cuda:0', grad_fn=<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(3.6857, 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(0.0994, 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(0.0019, 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(0.0014, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, 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(0.0035, 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(0.4374, 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(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, 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(0.0054, device='cuda:0', grad_fn=<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(0.0105, 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(0.0045, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7985, 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(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, 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(0.0060, 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(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0002, 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(0.0112, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, 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(0.0298, 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(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2538, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0022, 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(0.0021, 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(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0202, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9762, 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(0.0208, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, 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(0.0049, 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(0.0002, 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(0.0006, 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(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0115, 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(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.0415, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(0.0301, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, 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(0.2990, 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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3674, 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(0.0038, 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(0.0013, 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(0.4909, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6004, 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(0.0502, 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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, 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(0.0016, 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(0.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0194, 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(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, 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(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6158, device='cuda:0', grad_fn=<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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, 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(0.0618, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, 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(0.0004, 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(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, 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(0.0130, 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(0.6242, 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(2.1196, 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(0.0681, device='cuda:0', grad_fn=<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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, 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(0.8446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, 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(0.0004, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0189, 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(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0016, 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(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, 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(0.0042, 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(0.0028, 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(0.0003, 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(0.0192, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, 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(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0034, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, 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(0.0181, 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(0.0002, 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(0.0002, 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(0.0005, 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(0.0004, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6031, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<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(0.0006, 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(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7405, 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(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, 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(0.0026, 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(0.0034, 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(0.0007, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7921, 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(0.0031, 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(0.0033, 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(0.0002, 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(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, 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(0.0209, 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(0.0353, 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(0.0059, 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(0.0011, 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(0.2263, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, 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(0.0103, 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(0.0010, 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(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, 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(1.6189, 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(0.0015, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, 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(0.0146, 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(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, 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(0.0085, 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(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(0.0324, device='cuda:0', grad_fn=<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(0.0165, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, 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(0.0253, 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(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(6.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, 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(0.0004, 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(0.0003, 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(0.0009, 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(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, 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(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, 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(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, 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(0.0005, 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(0.0006, 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(0.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4650, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0157, 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(0.0013, 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>)
			------------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
				 {'precision': {'support': 0.9224343675417661, 'agreement': 0.5862068965517241, 'direct_attack': 0.22285714285714286, 'undercutter_attack': 0.44571428571428573, 'partial': 0.4583333333333333}, 'recall': {'support': 0.9246411483253588, 'agreement': 0.5795454545454546, 'direct_attack': 0.35454545454545455, 'undercutter_attack': 0.46706586826347307, 'partial': 0.1134020618556701}, 'f1': {'support': 0.9235364396654719, 'agreement': 0.5828571428571429, 'direct_attack': 0.27368421052631586, 'undercutter_attack': 0.456140350877193, 'partial': 0.18181818181818182}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7236652236652237, 'recall': 0.7236652236652237, 'f1': 0.7236652236652237, 'support': None}, 'macro_avg': {'precision': 0.5271092051996504, 'recall': 0.48783999750708223, 'f1': 0.48360726514886104, 'support': None}, 'weighted_avg': {'precision': 0.7342961398412156, 'recall': 0.7236652236652237, 'f1': 0.7204735830403617, 'support': None}}
Loss: tensor(0.0439, 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(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, 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(1.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, 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(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0151, 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(0.0289, 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(0.0005, 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(0.0887, 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(0.0015, 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(0.0147, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, 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(0.1464, 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(0.0144, 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(0.0565, 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(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, 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(0.0171, device='cuda:0', grad_fn=<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(0.0007, 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(0.0148, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<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(0.4229, device='cuda:0', grad_fn=<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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, 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(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, 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(0.0352, 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(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, 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(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, 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(0.0002, 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(0.0054, device='cuda:0', grad_fn=<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(0.1562, 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(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(0.0182, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0005, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.1489, 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(0.0004, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.3874, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, 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(0.0035, 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(0.0177, 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(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, 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(0.0004, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<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(0.0008, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, 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(0.0003, 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(0.0021, 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(0.0220, 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(0.0066, 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(0.0083, 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(0.0083, 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(0.0031, 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(0.0148, device='cuda:0', grad_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(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<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(0.0199, 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(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0010, 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(0.0022, 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(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(0.0027, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(0.0226, device='cuda:0', grad_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(0.0546, 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(0.0074, 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(0.0450, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, 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(0.5134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5830, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0003, 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(0.0003, 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(0.0303, 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(0.0016, 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(0.0022, 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(0.0125, 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(0.0178, 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(0.0121, 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(0.0115, 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(0.0004, 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(0.0747, 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(0.0423, 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(0.0034, 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(0.0005, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, 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(0.0005, 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(0.0046, 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(0.0007, 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(0.0160, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, 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(0.0022, 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(0.0157, 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(0.0048, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0003, 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(0.0006, 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(0.0175, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(0.0030, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0037, 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(0.0005, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3797, 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(0.0066, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(0.0035, 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(0.0076, 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(0.0003, 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(0.0018, 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(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0009, 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(0.0002, 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(0.0003, 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(0.0048, 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(0.0018, 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(0.0305, 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(0.0123, 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(0.0005, 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(0.0051, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0079, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_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(0.4348, 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(0.0041, 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(0.0009, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, 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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, 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(0.0320, 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(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(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(0.0002, 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(0.2726, 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(0.0003, 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(0.0029, 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(0.0006, 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(0.0053, 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(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, 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(0.1363, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, 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(0.0002, 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(0.0007, 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(0.0026, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0004, 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(0.0208, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0172, 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(0.0006, 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(0.0002, 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(0.0029, 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(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, 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(0.0002, 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(0.0057, 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(0.0155, 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(0.0064, 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(0.0015, 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(0.0020, 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(0.0144, device='cuda:0', grad_fn=<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(0.0112, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0005, 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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, 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(0.0004, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6296, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, 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(0.0098, 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(0.0003, 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(0.0003, 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(0.0020, 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(0.0606, 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(0.0007, 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(0.0002, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4094, 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(0.0004, 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(0.0090, 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(0.0043, 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(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(0.0113, 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(0.0033, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, 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(0.0036, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0005, 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(0.0046, 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(0.0012, 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(0.0002, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0036, 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(0.0003, 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(0.0042, 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(0.0214, 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(0.0002, 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(0.0089, 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>)
			------------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
				 {'precision': {'support': 0.8805646036916395, 'agreement': 0.6666666666666666, 'direct_attack': 0.2611464968152866, 'undercutter_attack': 0.4576271186440678, 'partial': 0.45454545454545453}, 'recall': {'support': 0.9700956937799043, 'agreement': 0.45454545454545453, 'direct_attack': 0.37272727272727274, 'undercutter_attack': 0.48502994011976047, 'partial': 0.05154639175257732}, 'f1': {'support': 0.9231644849174729, 'agreement': 0.5405405405405405, 'direct_attack': 0.3071161048689139, 'undercutter_attack': 0.4709302325581395, 'partial': 0.09259259259259259}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7344877344877345, 'recall': 0.7344877344877345, 'f1': 0.7344877344877345, 'support': None}, 'macro_avg': {'precision': 0.544110068072623, 'recall': 0.46678895058499387, 'f1': 0.46686879109553187, 'support': None}, 'weighted_avg': {'precision': 0.7234676007025209, 'recall': 0.7344877344877345, 'f1': 0.7130665558300244, 'support': None}}
Loss: tensor(0.0053, 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(0.0141, device='cuda:0', grad_fn=<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(0.0094, 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(0.0184, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.0393, 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(0.0104, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0069, 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(0.0005, 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(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0010, 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(0.0025, 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(0.0023, 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(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(0.0009, 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(0.0002, 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(0.0002, 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(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(0.0047, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, 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(0.0086, 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(0.0018, 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(0.0360, 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(0.0025, 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(0.0010, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0027, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0006, 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(0.0001, 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(0.0002, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, 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(0.0079, device='cuda:0', grad_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(0.0099, 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(0.0043, device='cuda:0', grad_fn=<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(0.0058, 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(0.0041, 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(0.0006, 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(0.0011, 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(0.0019, 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(0.0001, 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(0.0009, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0064, 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(0.0005, 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(0.0168, 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(0.0011, 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(0.0093, 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(0.0006, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, 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(0.0010, 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(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(0.0060, 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(0.0019, 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(0.0014, 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(0.0097, 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(0.0021, 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(0.0024, 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(0.0003, 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(0.0048, 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(0.0007, 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(0.0001, 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(0.0010, 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(0.0035, 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(0.0024, 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(0.0002, 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(0.0002, 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(0.0007, 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(0.0036, 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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2482, 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(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(0.0001, 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(0.0045, 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(0.0056, 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(0.0010, 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(0.0036, 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(0.0054, 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(0.0010, 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(0.0005, 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(0.0001, 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(0.0005, 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(0.0055, 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(0.0007, 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(0.0006, 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(0.0092, 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(0.0002, 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(0.0008, 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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, 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(0.0054, device='cuda:0', grad_fn=<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(0.0001, 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(0.0002, 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(0.0004, 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(0.0066, 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(0.2722, device='cuda:0', grad_fn=<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(0.0004, 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(0.0038, 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(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(0.0100, 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(0.0014, 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(0.0015, 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(0.0009, 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(0.0065, 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(0.0004, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, 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(0.0001, 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(0.0001, 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(0.0010, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0033, 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(0.0007, 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(0.0173, device='cuda:0', grad_fn=<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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0001, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, 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(0.0002, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0344e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1414e-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(0.0002, 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(0.0066, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0003, 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(0.0077, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6742e-05, device='cuda:0', grad_fn=<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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0001, 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(0.0009, 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(0.0108, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5746, 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(0.0001, 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(0.0070, device='cuda:0', grad_fn=<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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, 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(9.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0015, 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(0.0034, 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(0.0002, 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(0.0002, 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(0.0074, 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(0.0046, 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(0.0021, 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(0.0020, 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(0.0008, 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(0.0008, 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(0.0101, 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(9.8154e-05, 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(0.0002, 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(0.0021, 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(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0020, 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(0.0075, 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(0.0001, 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(0.0093, 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(0.0009, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0196, 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(0.0027, 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(0.0001, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0014, 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(0.0176, 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(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, 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(0.0009, 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(0.0003, 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(0.0075, 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(0.0067, 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(0.0001, 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(0.0040, 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(0.0021, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0007, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0175, 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(0.1237, 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(0.0002, 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(0.1141, 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(0.0003, 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(0.0005, 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(0.0102, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7822e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, 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(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0147, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, 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(0.0001, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0008, 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(8.9285e-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(0.0051, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-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(0.0067, 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(0.0018, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, 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(0.0008, 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(0.0015, 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(0.0030, 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(0.0004, 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(0.0090, 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(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(0.0003, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0186, 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(0.0010, 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(0.0016, 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(0.0001, 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(0.0005, 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(8.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1535e-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(0.0091, 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(9.5552e-05, 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(0.0013, 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(0.0001, 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(0.0019, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7126e-05, device='cuda:0', grad_fn=<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(0.0001, 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(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(0.0178, 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(0.0007, 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(0.0049, 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(0.0023, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0026, 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(0.0009, 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(0.0030, 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
				 {'precision': {'support': 0.8705882352941177, 'agreement': 0.6759259259259259, 'direct_attack': 0.25, 'undercutter_attack': 0.4727272727272727, 'partial': 0.45454545454545453}, 'recall': {'support': 0.9736842105263158, 'agreement': 0.4147727272727273, 'direct_attack': 0.35454545454545455, 'undercutter_attack': 0.46706586826347307, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9192546583850931, 'agreement': 0.5140845070422536, 'direct_attack': 0.29323308270676696, 'undercutter_attack': 0.4698795180722891, 'partial': 0.1680672268907563}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7316017316017316, 'recall': 0.7316017316017316, 'f1': 0.7316017316017315, 'support': None}, 'macro_avg': {'precision': 0.5447573776985541, 'recall': 0.462632208822625, 'f1': 0.4729037986194318, 'support': None}, 'weighted_avg': {'precision': 0.7195606719373803, 'recall': 0.7316017316017316, 'f1': 0.7114024583503569, 'support': None}}
Loss: tensor(0.0067, 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(0.0015, 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(0.0180, 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(0.0021, 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(0.0155, 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(0.0043, 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(0.0037, device='cuda:0', grad_fn=<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(0.0028, device='cuda:0', grad_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(0.0001, 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(0.0049, 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(0.0019, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7176e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0028, 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(0.0003, 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(0.0105, 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(0.0006, 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(0.0005, 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(0.0001, 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(0.0053, 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(0.0010, 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(0.0007, 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(0.0006, 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(0.0006, 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(0.0071, 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(0.0035, 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(0.0036, 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(0.0002, 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(0.0025, 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(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(0.0074, 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(0.0001, 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(0.0018, 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(0.0013, 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(0.0023, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<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(0.0001, 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(9.4462e-05, device='cuda:0', grad_fn=<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(0.0029, 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(0.0001, 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(0.0011, 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(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(0.0012, 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(0.0034, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0002, 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(7.8961e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4462e-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(0.0068, 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(0.0002, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6672e-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(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(0.0009, 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(0.0001, 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(0.0016, 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(0.0019, 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(0.0045, 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(0.0002, 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(0.0026, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8226e-05, 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(0.0003, 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(0.0010, 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(0.0015, 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(0.0016, 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(0.0004, 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(0.0001, 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(0.0011, 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(0.0015, 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(0.0070, 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(0.0022, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0008, 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(0.0010, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4300e-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(0.0019, 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(0.0073, 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(0.0015, 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(0.0002, 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(0.0005, 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(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, 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(0.0053, 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(0.0005, 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(0.0015, 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(0.0016, 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(0.0049, 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(0.0041, device='cuda:0', grad_fn=<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(0.0008, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2029e-05, device='cuda:0', grad_fn=<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(0.0039, 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(0.0015, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0799e-05, 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(9.8609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0014, 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(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(0.0028, 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(0.0002, 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(0.0024, 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(0.0009, 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(0.0114, device='cuda:0', grad_fn=<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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2653e-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(0.0020, 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(0.0012, 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(0.0028, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0017, 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(0.0071, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4099e-05, 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(0.0012, 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(0.0011, 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(0.0018, 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(0.0028, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3079e-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(0.0002, 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(0.0002, 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(0.0029, 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(6.1705e-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(0.0001, 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(0.0034, device='cuda:0', grad_fn=<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(7.8901e-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(0.0011, 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(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(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(0.0007, 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(0.0035, 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(0.0015, 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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0020, 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(0.0021, 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(0.0014, 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(0.0011, 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(0.0004, 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(0.0002, 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(0.0012, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0324e-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(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0051, 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(0.0087, 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(0.0011, 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(0.0001, 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(0.0002, 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(0.0008, 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(0.0006, 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(0.0018, 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(0.0034, 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(0.0024, 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(0.0001, 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(0.0024, 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(9.1736e-05, 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(5.7679e-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(0.0013, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5824e-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(0.0008, 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(0.0048, device='cuda:0', grad_fn=<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(0.0001, 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(0.0007, 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(0.0007, 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(0.0013, 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(0.0001, 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(0.0029, 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(0.0047, 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(0.0012, 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(0.0001, 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(0.0022, 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(0.0030, 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(0.0097, 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(0.0002, 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(9.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1262, 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(6.5187e-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(0.0013, 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(0.0021, 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(0.0025, 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(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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0708e-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(0.0033, 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(0.0023, 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(0.0037, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, 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(0.0034, 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(0.0002, 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(7.7751e-05, 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(0.0024, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9536e-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(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(7.2604e-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(0.0011, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9679e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0042, device='cuda:0', grad_fn=<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(0.0017, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, 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(0.0008, 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(0.0023, 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(0.0209, 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(9.7459e-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(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(8.3654e-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(0.0040, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0002, 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(0.0083, 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(0.0040, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3250e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0027, 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(0.0013, 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(0.0002, 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(0.0015, 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(0.0006, 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(0.0001, 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(0.0012, 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(0.0004, 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(0.0009, 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(0.0006, 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(0.0012, 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(0.0005, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9708e-05, 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(0.0001, 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(0.0004, 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(0.0017, 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(0.0040, 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(0.0020, 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(8.9648e-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(5.0957e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7731e-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(0.0002, 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(0.0002, 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(0.0007, 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(0.0115, 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(9.0678e-05, 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(0.0002, 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(0.0066, 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(9.3978e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7448e-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(0.0052, 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>)
			------------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
				 {'precision': {'support': 0.8800436205016358, 'agreement': 0.6201550387596899, 'direct_attack': 0.23943661971830985, 'undercutter_attack': 0.47023809523809523, 'partial': 0.3333333333333333}, 'recall': {'support': 0.965311004784689, 'agreement': 0.45454545454545453, 'direct_attack': 0.3090909090909091, 'undercutter_attack': 0.47305389221556887, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9207073588134627, 'agreement': 0.5245901639344261, 'direct_attack': 0.2698412698412698, 'undercutter_attack': 0.47164179104477616, 'partial': 0.1574803149606299}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7287157287157288, 'recall': 0.7287157287157288, 'f1': 0.7287157287157288, 'support': None}, 'macro_avg': {'precision': 0.5086413415102128, 'recall': 0.4610188088283552, 'f1': 0.46885217971891296, 'support': None}, 'weighted_avg': {'precision': 0.7085605172930609, 'recall': 0.7287157287157288, 'f1': 0.7112276552371661, 'support': None}}
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(0.0007, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8951e-05, 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(0.0003, 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(0.0047, 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(0.0066, 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(5.6013e-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(7.4511e-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(0.0011, 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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4281e-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(0.0002, 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(0.0005, 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(6.2129e-05, 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(0.0006, 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(0.0005, 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(0.0001, 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(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(0.0011, 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(7.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5774e-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(0.0069, 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(0.0009, 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(4.9474e-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(0.0017, 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(0.0009, 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(0.0011, 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(0.0001, 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(9.3917e-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(0.0020, 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(0.0014, 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(0.0001, 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(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(0.0004, 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(0.0017, 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(0.0001, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-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(0.0022, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4845e-05, device='cuda:0', grad_fn=<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(0.0001, 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(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(0.0025, 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(0.0019, 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(0.0002, 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(0.0028, 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(0.0003, 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(0.0015, 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(0.0007, 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(0.0007, 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(0.0002, 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(6.9698e-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(0.0033, 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(0.0004, 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(0.0017, 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(8.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4886e-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(9.6006e-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(0.0011, 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(6.2401e-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(0.0033, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-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(0.0009, 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(0.0008, 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(0.0007, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0142e-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(0.0015, 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(0.0018, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-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(0.0032, 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(0.0019, 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(0.0006, 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(0.0019, 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(0.0011, 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(0.0001, 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(0.0001, 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(9.8397e-05, 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(0.0061, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, 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(7.8326e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8689e-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(0.0011, 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(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(0.0002, 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(0.0016, 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(0.0034, 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(0.0005, 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(0.0024, 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(7.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1232e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0022, 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(0.0015, 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(0.0010, 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(0.0003, 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(0.0017, 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(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(0.0039, 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(0.0005, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4944e-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(0.0026, 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(0.0002, 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(0.0004, 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(0.0009, 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(0.0012, 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(0.0018, 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(0.0028, 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(0.0008, 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(0.0007, 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(6.3521e-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(0.0003, 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(7.6661e-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(0.0008, 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(0.0002, 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(0.0009, 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(8.9195e-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(0.0001, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-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(0.0031, 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(0.0005, 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(9.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-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(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(0.0025, 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(0.0017, 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(0.0003, 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(0.0003, 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(0.0002, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<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(4.6628e-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(0.0026, 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(0.0002, 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(0.0070, 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(0.0024, 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(0.0021, 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(0.0010, 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(0.0034, 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(0.0015, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9151e-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(8.2898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8276e-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(0.0001, 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(0.0006, 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(0.0002, 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(0.0015, 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(0.0023, 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(0.0014, 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(0.0020, 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(0.0010, 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(0.0003, 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(0.0044, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8044e-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(0.0012, 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(0.0001, 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(0.0002, 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(0.0074, device='cuda:0', grad_fn=<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(0.0010, 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(0.0004, 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(8.7196e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9124e-05, device='cuda:0', grad_fn=<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(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(0.0008, 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(0.0086, 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(0.0002, 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(9.9669e-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(0.0031, 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(0.0003, 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(0.0002, 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(4.6113e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8770e-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(0.0009, 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(9.3221e-05, device='cuda:0', grad_fn=<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(0.0009, 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(0.0005, 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(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(0.0018, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2383e-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(0.0017, 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(0.0002, 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(0.0013, 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(0.0001, 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(0.0008, 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(0.0002, 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(9.6852e-05, 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(6.0857e-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(0.0049, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9618e-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(5.9888e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0100, 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(0.0001, 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(0.0028, 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(0.0031, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3675e-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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0254e-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(0.0033, 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(0.0047, device='cuda:0', grad_fn=<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(0.0038, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9537e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, 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(0.0014, 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(6.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5066e-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(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6520e-05, 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(0.0006, 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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4714e-05, 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(0.0025, 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(7.2544e-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(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(0.0014, 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(0.0004, 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(0.0031, 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(0.0009, 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(0.0010, 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(9.6399e-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(6.8971e-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(0.0020, 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(0.0021, 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(6.8396e-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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7308e-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(7.3905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6377e-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(0.0003, 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(0.0040, 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(0.0009, 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(0.0047, 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(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(0.0008, 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(0.0042, 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(0.0061, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7751e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, 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(9.4462e-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(8.7893e-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(6.5368e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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
				 {'precision': {'support': 0.8775731310942578, 'agreement': 0.6422764227642277, 'direct_attack': 0.24113475177304963, 'undercutter_attack': 0.47337278106508873, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9688995215311005, 'agreement': 0.44886363636363635, 'direct_attack': 0.3090909090909091, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9209778283115406, 'agreement': 0.5284280936454849, 'direct_attack': 0.2709163346613545, 'undercutter_attack': 0.4761904761904762, 'partial': 0.1574803149606299}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7308802308802309, 'recall': 0.7308802308802309, 'f1': 0.7308802308802309, 'support': None}, 'macro_avg': {'precision': 0.5135380840059914, 'recall': 0.4617977533316931, 'f1': 0.47079860955389724, 'support': None}, 'weighted_avg': {'precision': 0.7103919180862498, 'recall': 0.7308802308802309, 'f1': 0.7125115482235159, 'support': None}}
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(0.0021, 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(0.0003, 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(0.0002, 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(0.0009, 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(0.0023, 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(0.0015, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1405e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8104e-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(0.0016, 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(0.0028, 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(0.0001, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-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(7.2785e-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(0.0014, 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(0.0028, 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(0.0006, 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(0.0003, 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(0.0015, 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(8.0566e-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(0.0031, 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(8.0506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-05, device='cuda:0', grad_fn=<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(0.0041, 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(0.0002, 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(0.0007, 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(0.0002, 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(7.0545e-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(0.0011, 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(6.3279e-05, device='cuda:0', grad_fn=<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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4459e-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(0.0036, 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(0.0016, 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(0.0002, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8770e-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(0.0005, 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(0.0002, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0007, 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(0.0007, 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(0.0008, 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(0.0006, 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(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0017, 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(0.0001, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4795e-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(0.0006, 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(0.0017, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-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(0.0001, 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(0.0006, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-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(0.0001, 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(0.0019, 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(9.1192e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0627e-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(0.0009, 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(0.0024, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9335e-05, 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(0.0001, 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(0.0107, device='cuda:0', grad_fn=<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(0.0023, 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(0.0002, 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(0.0021, 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(0.0020, 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(0.0017, 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(0.0002, device='cuda:0', grad_fn=<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(0.0010, 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(0.0027, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-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(0.0006, 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(8.6135e-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(8.3562e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2997e-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(0.0011, 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(0.0002, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0015, 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(0.0036, 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(0.0024, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2240e-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(0.0001, 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(8.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-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(0.0006, 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(0.0008, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7336e-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(6.0040e-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(6.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-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(0.0035, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(7.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1695e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0008, 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(0.0017, 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(0.0027, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-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(9.0012e-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(0.0001, 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(0.0041, device='cuda:0', grad_fn=<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(0.0004, 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(0.0009, 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(0.0064, 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(0.0027, 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(0.0007, 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(0.0009, 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(0.0020, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5552e-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(0.0014, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0103e-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(0.0001, 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(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9678e-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(0.0009, 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(6.9364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0112e-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(8.0445e-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(6.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5428e-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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-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(0.0014, 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(8.1352e-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(0.0001, 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(0.0018, 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(0.0011, 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(0.0001, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0018, 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(9.3643e-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(0.0025, 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(0.0002, 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(0.0001, 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(0.0013, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0323e-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(0.0022, 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(0.0010, 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(0.0011, 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(0.0024, 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(0.0069, 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(5.6558e-05, device='cuda:0', grad_fn=<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(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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1553e-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(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(0.0020, 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(0.0007, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3855e-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(5.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6711e-05, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4765e-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(0.0016, 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(0.0010, 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(0.0003, 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(4.5144e-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(7.2331e-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(7.5178e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4641e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<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(9.2432e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-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(0.0007, 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(0.0004, 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(7.6630e-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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7669e-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(0.0040, 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(0.0006, 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(0.0008, 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(0.0007, 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(0.0017, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4561e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-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(0.0018, 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(0.0036, 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(9.5491e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8891e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8710e-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(0.0014, 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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5380e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4902e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4512e-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(0.0041, 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(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5400e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, 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(0.0025, 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(4.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6006e-05, 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(7.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-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(8.4622e-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(0.0003, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-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(0.0008, 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(0.0002, 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(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(0.0011, 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(0.0006, 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(6.7942e-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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-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(0.0001, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9768e-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(0.0022, 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(7.0969e-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(8.0384e-05, 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(0.0002, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7408e-05, 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(0.0015, 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(0.0016, 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(0.0002, 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(9.4308e-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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4683e-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(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(0.0015, 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(0.0033, device='cuda:0', grad_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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5087e-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(0.0017, 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
				 {'precision': {'support': 0.8766233766233766, 'agreement': 0.6475409836065574, 'direct_attack': 0.2446043165467626, 'undercutter_attack': 0.47058823529411764, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9688995215311005, 'agreement': 0.44886363636363635, 'direct_attack': 0.3090909090909091, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9204545454545454, 'agreement': 0.5302013422818792, 'direct_attack': 0.2730923694779116, 'undercutter_attack': 0.4747774480712166, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7308802308802309, 'recall': 0.7308802308802309, 'f1': 0.7308802308802309, 'support': None}, 'macro_avg': {'precision': 0.5123875114464209, 'recall': 0.4617977533316931, 'f1': 0.47095514105711056, 'support': None}, 'weighted_avg': {'precision': 0.7096748835979825, 'recall': 0.7308802308802309, 'f1': 0.7123374319711935, 'support': None}}
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(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(0.0021, 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(0.0028, 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(0.0002, 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(0.0005, 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(0.0004, 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(0.0010, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1514e-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(0.0028, 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(0.0004, 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(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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-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(5.6528e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-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(7.4965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-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(0.0094, 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(7.9477e-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(9.4189e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0373e-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(6.4581e-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(0.0003, 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(0.0026, 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(8.9133e-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(0.0013, 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(5.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, 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(0.0014, 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(8.6440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7730e-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(0.0013, 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(0.0019, 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(0.0018, 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(8.2203e-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(6.7427e-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(4.7930e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2504e-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(0.0003, 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(0.0011, 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(5.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-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(0.0014, 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(0.0004, 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(0.0003, 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(0.0005, 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(0.0032, 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(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6288e-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(0.0008, 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(5.5589e-05, device='cuda:0', grad_fn=<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(6.4551e-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(0.0002, 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(0.0007, 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(0.0020, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3553e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<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(6.5035e-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(8.4530e-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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-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(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-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(0.0016, 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(0.0010, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2522e-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(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, 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(0.0003, 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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2804e-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(6.4187e-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(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7054e-05, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-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(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, 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(0.0013, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3452e-05, 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(5.4590e-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(0.0001, 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(6.2190e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0007, 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(0.0031, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3996e-05, 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(5.5377e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-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(0.0002, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1433e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0006, 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(0.0013, 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(0.0020, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6803e-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(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-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(0.0018, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0017, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, 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(0.0004, 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(0.0002, 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(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(7.9957e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3290e-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(0.0030, 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(0.0026, 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(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(0.0002, 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(0.0001, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8720e-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(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, 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(0.0009, 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(0.0011, 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(0.0010, 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(4.4992e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, 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(0.0015, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6175e-05, device='cuda:0', grad_fn=<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(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5427e-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(7.2634e-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(8.8224e-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(5.8495e-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(8.6325e-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(0.0002, 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(0.0031, 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(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6610e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, 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(5.4469e-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(0.0015, 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(0.0009, 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(0.0002, 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(8.6863e-05, 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(0.0001, 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(5.4196e-05, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0003, 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(0.0020, 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(0.0029, 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(8.6561e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3503e-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(0.0025, 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(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-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(0.0002, 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(0.0015, 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(7.6600e-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(0.0006, 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(6.0585e-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(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-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(0.0009, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-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(5.7800e-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(9.8396e-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(0.0004, 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(0.0011, 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(0.0029, 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(0.0003, 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(9.0678e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1373e-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(5.9071e-05, device='cuda:0', grad_fn=<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(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4632e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8305e-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(0.0008, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-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(0.0008, 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(0.0008, 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(0.0029, 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(9.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7045e-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(7.1908e-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(4.1299e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-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(0.0028, 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(5.9313e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4733e-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(0.0009, 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(7.4511e-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(5.8526e-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(0.0006, 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(0.0012, 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(7.1242e-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(0.0001, 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(0.0010, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1784e-05, device='cuda:0', grad_fn=<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(5.6134e-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(7.6812e-05, device='cuda:0', grad_fn=<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(4.9353e-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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-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(0.0011, 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(0.0001, 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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6085e-05, 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(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2785e-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(7.0545e-05, device='cuda:0', grad_fn=<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(0.0009, 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(7.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-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(0.0006, 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(6.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7700e-05, 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(6.5641e-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(0.0005, 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(0.0026, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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
				 {'precision': {'support': 0.8766233766233766, 'agreement': 0.6475409836065574, 'direct_attack': 0.2463768115942029, 'undercutter_attack': 0.47058823529411764, 'partial': 0.3125}, 'recall': {'support': 0.9688995215311005, 'agreement': 0.44886363636363635, 'direct_attack': 0.3090909090909091, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9204545454545454, 'agreement': 0.5302013422818792, 'direct_attack': 0.2741935483870968, 'undercutter_attack': 0.4747774480712166, 'partial': 0.15503875968992248}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7308802308802309, 'recall': 0.7308802308802309, 'f1': 0.7308802308802309, 'support': None}, 'macro_avg': {'precision': 0.5107258814236509, 'recall': 0.4617977533316931, 'f1': 0.4709331287769321, 'support': None}, 'weighted_avg': {'precision': 0.7091100581106616, 'recall': 0.7308802308802309, 'f1': 0.7123400577792259, 'support': None}}
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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0513e-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(8.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-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(0.0012, 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(6.9789e-05, device='cuda:0', grad_fn=<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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7681e-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(0.0017, device='cuda:0', grad_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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<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(8.3624e-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(8.1595e-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(9.9881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-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(0.0001, 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(0.0015, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1707e-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(0.0001, device='cuda:0', grad_fn=<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(9.3039e-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(9.1676e-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(0.0008, 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(9.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4239e-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(0.0018, 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(0.0040, device='cuda:0', grad_fn=<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(0.0008, 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(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1232e-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(6.3370e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0729e-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(7.5782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9819e-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(9.6914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-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(7.4117e-05, device='cuda:0', grad_fn=<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(0.0011, 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(6.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9937e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0798e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5409e-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(0.0001, 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(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1130e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0003, 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(7.2059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-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(6.9515e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3281e-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(7.9960e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6669e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3662e-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(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9798e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-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(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2413e-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(7.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3915e-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(8.6044e-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(0.0027, 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(2.9219e-05, 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(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-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(0.0004, 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(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-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(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9596e-05, 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(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7196e-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(0.0011, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7670e-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(0.0016, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6228e-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(0.0017, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-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(0.0006, 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(0.0002, 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(7.1937e-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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-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(0.0004, 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(9.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0006, 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(6.1160e-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(5.8617e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6076e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6024e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5380e-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(9.9813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0042, 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(0.0002, 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(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-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(0.0013, 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(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-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(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3796e-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(7.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4805e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0015, device='cuda:0', grad_fn=<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(9.3068e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-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(0.0020, 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(6.1160e-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(0.0039, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0007, 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(7.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3752e-05, 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(6.9032e-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(6.5973e-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(9.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3190e-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(4.9776e-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(0.0004, 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(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-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(8.7166e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0010, 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(0.0020, 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(0.0028, 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(0.0008, 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(0.0029, 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(6.3461e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-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(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(0.0016, 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(8.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1256e-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(5.9524e-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(0.0001, 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(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(7.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-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(9.5830e-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(3.7726e-05, 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(7.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0083e-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(8.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8426e-05, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-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(0.0041, 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(6.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2229e-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(4.8959e-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(0.0014, 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(0.0008, 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(0.0005, 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(9.2582e-05, 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(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(9.6572e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-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(0.0013, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0172e-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(0.0013, 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(7.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8336e-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(0.0012, 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(6.8063e-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(0.0009, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-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(0.0007, 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(0.0011, 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(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(0.0020, 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(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-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(0.0002, 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(4.6688e-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(0.0001, 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(0.0017, 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(8.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3391e-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(5.5075e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5107e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-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(0.0005, 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(0.0005, 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(8.5983e-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(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(0.0001, 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(0.0014, 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(0.0003, 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(0.0002, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4864e-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(5.4135e-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(0.0022, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4333e-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(0.0011, 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(0.0012, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-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(0.0004, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-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(7.8356e-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(6.2704e-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>)
			------------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
				 {'precision': {'support': 0.8766233766233766, 'agreement': 0.6475409836065574, 'direct_attack': 0.2391304347826087, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9688995215311005, 'agreement': 0.44886363636363635, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9204545454545454, 'agreement': 0.5302013422818792, 'direct_attack': 0.2661290322580645, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7301587301587301, 'recall': 0.7301587301587301, 'f1': 0.7301587301587301, 'support': None}, 'macro_avg': {'precision': 0.510742339496755, 'recall': 0.4599795715135112, 'f1': 0.46928154021191554, 'support': None}, 'weighted_avg': {'precision': 0.7089088610024454, 'recall': 0.7301587301587301, 'f1': 0.7116155369609435, '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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6161e-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(0.0006, 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(4.5295e-05, device='cuda:0', grad_fn=<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(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1167e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3776e-05, 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(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4642e-05, device='cuda:0', grad_fn=<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(0.0015, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-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(8.7255e-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(5.2804e-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(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<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(0.0006, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4520e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-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(0.0006, 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(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4118e-05, 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(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(9.4699e-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(5.3864e-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(0.0004, 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(6.9153e-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(0.0011, 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(0.0022, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7275e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4743e-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(0.0014, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5002e-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(0.0001, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-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(0.0007, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-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(0.0021, 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(6.4944e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3935e-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(0.0012, 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(6.9274e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0545e-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(0.0005, 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(0.0002, 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(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8255e-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(8.0891e-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(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-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(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(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6388e-05, 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(0.0012, 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(9.4885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, 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(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-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(0.0011, 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(0.0004, 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(0.0015, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8041e-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(0.0007, 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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-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(0.0002, 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(0.0004, 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(9.0799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3463e-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(0.0004, 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(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3855e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-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(0.0009, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1929e-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(5.1310e-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(6.9940e-05, device='cuda:0', grad_fn=<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(4.5417e-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(0.0004, 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(0.0013, 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(0.0014, 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(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(3.8937e-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(0.0007, 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(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7518e-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(0.0005, 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(5.6891e-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(0.0024, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-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(0.0020, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-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(0.0006, 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(0.0009, 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(0.0008, 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(8.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0253e-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(0.0004, 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(8.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1757e-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(7.5389e-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(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-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(0.0004, 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(7.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(8.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-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(0.0007, 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(6.5550e-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(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6216e-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(3.6636e-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(0.0003, 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(0.0005, 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(0.0001, 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(0.0010, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8941e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9553e-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(3.5092e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6547e-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(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5923e-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(5.0503e-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(0.0008, 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(0.0009, 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(0.0001, 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(5.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6386e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-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(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9746e-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(6.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5527e-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(7.1968e-05, 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(0.0011, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, device='cuda:0', grad_fn=<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(0.0041, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-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(9.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-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(0.0028, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-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(0.0007, 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(0.0002, 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(9.8030e-05, 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(0.0006, 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(6.8305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6748e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5166e-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(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(5.4288e-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(0.0004, 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(5.8495e-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(0.0017, 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(0.0016, 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(4.8929e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(7.2694e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4637e-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(0.0007, 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(0.0004, 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(0.0014, 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(5.9525e-05, 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(0.0003, 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(0.0008, 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(0.0009, 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(5.3773e-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(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(9.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4238e-05, device='cuda:0', grad_fn=<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(5.9010e-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(3.2822e-05, 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(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2493e-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(6.1674e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-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(0.0002, device='cuda:0', grad_fn=<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(0.0010, 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(9.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-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(4.6991e-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(0.0020, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-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(0.0004, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4308e-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(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1927e-05, device='cuda:0', grad_fn=<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(8.6983e-05, 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(0.0011, 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(0.0016, 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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3017e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8775731310942578, 'agreement': 0.6504065040650406, 'direct_attack': 0.2391304347826087, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9688995215311005, 'agreement': 0.45454545454545453, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9209778283115406, 'agreement': 0.5351170568561872, 'direct_attack': 0.2661290322580645, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7308802308802309, 'recall': 0.7308802308802309, 'f1': 0.7308802308802309, 'support': None}, 'macro_avg': {'precision': 0.5115053944826278, 'recall': 0.4611159351498749, 'f1': 0.47036933969817624, 'support': None}, 'weighted_avg': {'precision': 0.7098456043923081, 'recall': 0.7308802308802309, 'f1': 0.7125553856142812, 'support': None}}
Loss: tensor(7.9900e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-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(0.0006, 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(0.0009, 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(8.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-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(4.7112e-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(0.0014, 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(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-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(0.0023, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6357e-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(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(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(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5883e-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(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(7.3906e-05, device='cuda:0', grad_fn=<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(5.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-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(0.0034, 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(0.0004, 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(0.0010, 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(0.0001, 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(9.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-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(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8316e-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(7.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-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(7.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2220e-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(5.3379e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0001, 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(0.0001, 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(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7709e-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(0.0003, 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(3.1338e-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(6.9240e-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(6.2401e-05, 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(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0004, 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(4.0936e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5032e-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(0.0016, 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(3.3972e-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(8.0748e-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(4.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6529e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-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(2.4404e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9556e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<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(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, 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(0.0001, 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(7.7627e-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(4.1208e-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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-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(4.1814e-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(8.3047e-05, 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(0.0001, 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(0.0007, 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(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9962e-05, device='cuda:0', grad_fn=<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(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4018e-05, 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(4.8263e-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(5.0624e-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(7.2240e-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(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0385e-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(0.0001, 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(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(6.8032e-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(0.0008, 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(4.5538e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-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(8.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8658e-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(8.9476e-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(9.7608e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0007, 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(0.0002, 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(6.6639e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1544e-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(0.0004, 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(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9171e-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(8.6874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, 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(0.0008, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0928e-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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-05, 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(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-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(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8342e-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(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-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(5.6194e-05, 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(7.0473e-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(0.0003, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2996e-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(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(0.0010, 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(0.0019, 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(9.0161e-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(4.2904e-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(0.0017, 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(0.0009, 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(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1079e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-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(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-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(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0082e-05, device='cuda:0', grad_fn=<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(6.1493e-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(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9726e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5610e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-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(0.0006, 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(0.0001, 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(4.7415e-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(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1504e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-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(3.0914e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<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(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8435e-05, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8735e-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(0.0008, 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(0.0006, 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(0.0005, 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(3.9240e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1412e-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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8083e-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(0.0005, 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(4.8323e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0015, 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(7.3329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-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(8.3066e-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(5.0927e-05, 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(0.0008, 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(4.9565e-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(7.6570e-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(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9133e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9827e-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(7.1624e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-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(4.5174e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-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(3.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, 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(0.0005, 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(5.2320e-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(7.7554e-05, 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(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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, 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(4.7475e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-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(0.0005, 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(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-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(0.0006, 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(4.2207e-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(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-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(0.0011, 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(5.9071e-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(6.4429e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-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(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, 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(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(0.0014, 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(3.9301e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-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(2.9582e-05, 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(4.7597e-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(6.4127e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, 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(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9321e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-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(0.0011, 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(3.3790e-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(0.0010, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-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(0.0011, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7467e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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>)
			------------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
				 {'precision': {'support': 0.8786565547128927, 'agreement': 0.648, 'direct_attack': 0.2426470588235294, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9700956937799043, 'agreement': 0.4602272727272727, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9221148379761227, 'agreement': 0.5382059800664453, 'direct_attack': 0.26829268292682923, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7323232323232324, 'recall': 0.7323232323232324, 'f1': 0.7323232323232324, 'support': None}, 'macro_avg': {'precision': 0.5119441032015308, 'recall': 0.46249153323599923, 'f1': 0.47164725640689725, 'support': None}, 'weighted_avg': {'precision': 0.71047260733187, 'recall': 0.7323232323232324, 'f1': 0.7138051634917417, 'support': None}}
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(0.0004, 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(4.6779e-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(0.0003, 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(9.9488e-05, device='cuda:0', grad_fn=<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(7.7448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-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(0.0006, 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(0.0007, 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(0.0003, 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(6.3340e-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(7.3632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5888e-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(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9748e-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(6.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<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(5.8223e-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(8.6257e-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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-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(0.0001, 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(0.0002, 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(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-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(0.0002, 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(4.8868e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3975e-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(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-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(5.8710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6206e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9432e-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(0.0009, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-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(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-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(6.7202e-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(4.4222e-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(3.8059e-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(9.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(5.6104e-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(0.0002, 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(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3024e-05, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399e-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(3.3215e-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(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-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(4.9504e-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(0.0003, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1219e-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(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4908e-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(0.0006, 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(3.1308e-05, device='cuda:0', grad_fn=<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(0.0028, 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(0.0001, 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(0.0001, 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(0.0006, 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(5.6346e-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(4.4902e-05, 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(8.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-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(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-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(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(0.0010, 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(0.0011, 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(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(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7747e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9569e-05, 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(7.3210e-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(6.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3542e-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(3.6091e-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(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, 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(5.2561e-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(0.0015, 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(0.0006, 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(0.0002, 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(0.0002, 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(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-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(0.0007, 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(7.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-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(0.0011, 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(3.9543e-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(6.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9858e-05, 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(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(5.6619e-05, 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(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(6.3854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-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(8.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6184e-05, device='cuda:0', grad_fn=<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(0.0015, 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(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(4.0724e-05, 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(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(0.0004, 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(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(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(0.0005, 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(3.1429e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<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(4.7354e-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(9.2001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(7.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3560e-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(5.6225e-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(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(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0029e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-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(0.0003, 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(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-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(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0003, 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(0.0008, 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(0.0010, 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(0.0005, 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(4.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-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(0.0008, 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(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-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(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-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(5.6376e-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(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7130e-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(3.9634e-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(0.0003, 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(8.8890e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-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(7.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0006, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-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(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-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(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-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(0.0007, 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(0.0010, 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(2.6463e-05, 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(4.0996e-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(5.1660e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-05, device='cuda:0', grad_fn=<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(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-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(0.0014, 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(0.0003, 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(0.0003, 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(0.0005, 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(3.6303e-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(0.0002, 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(4.2540e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8092e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9143e-05, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6519e-05, 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(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, 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(3.6394e-05, device='cuda:0', grad_fn=<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(0.0005, 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(9.1102e-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(0.0001, 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(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(6.2340e-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(0.0006, 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(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3047e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9637e-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(0.0003, 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(7.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-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(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-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(6.2880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-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(2.5918e-05, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8747e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.2116e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-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(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-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(8.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(9.0400e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1719e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3106e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-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
				 {'precision': {'support': 0.8796095444685467, 'agreement': 0.6507936507936508, 'direct_attack': 0.2426470588235294, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9700956937799043, 'agreement': 0.4659090909090909, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9226393629124006, 'agreement': 0.543046357615894, 'direct_attack': 0.26829268292682923, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.733044733044733, 'recall': 0.733044733044733, 'f1': 0.733044733044733, 'support': None}, 'macro_avg': {'precision': 0.5126934313113918, 'recall': 0.46362789687236294, 'f1': 0.4727202369040425, 'support': None}, 'weighted_avg': {'precision': 0.7114021758566964, 'recall': 0.733044733044733, 'f1': 0.7147361947294265, 'support': None}}
Loss: tensor(7.4114e-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(5.2431e-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(5.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, 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(6.6549e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7156e-05, device='cuda:0', grad_fn=<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(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8143e-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(7.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-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(3.4002e-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(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(2.9612e-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(5.0594e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3652e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6014e-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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0019, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-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(4.3683e-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(0.0008, 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(0.0009, 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(3.1187e-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(7.5056e-05, device='cuda:0', grad_fn=<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(0.0006, 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(7.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3457e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4462e-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(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, 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(9.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-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(3.9664e-05, device='cuda:0', grad_fn=<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(3.7060e-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(0.0003, 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(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-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(0.0016, 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(0.0005, 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(9.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-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(0.0004, 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(0.0023, 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(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4179e-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(4.0421e-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(0.0006, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-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(7.0091e-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(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9151e-05, device='cuda:0', grad_fn=<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(4.7536e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8058e-05, device='cuda:0', grad_fn=<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(0.0012, 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(0.0003, 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(7.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-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(9.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-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(2.3072e-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(8.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-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(0.0004, 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(0.0004, 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(0.0019, device='cuda:0', grad_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(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-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(2.2618e-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(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5680e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6134e-05, 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(2.7432e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, 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(4.1541e-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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<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(3.0672e-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(6.9424e-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(6.5160e-05, device='cuda:0', grad_fn=<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(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6165e-05, device='cuda:0', grad_fn=<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(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9635e-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(0.0005, 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(0.0010, 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(7.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-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(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-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(6.2613e-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(5.3955e-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(0.0001, 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(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, 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(2.4707e-05, device='cuda:0', grad_fn=<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(3.2095e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(3.9997e-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(6.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-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>)
Loss: tensor(9.0162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2716e-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(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1481e-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(0.0007, 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(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-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(7.9839e-05, 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(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7174e-05, device='cuda:0', grad_fn=<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(9.7019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, 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(3.3336e-05, 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(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4501e-05, device='cuda:0', grad_fn=<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(4.2813e-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(8.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-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(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2734e-05, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<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(4.6991e-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(9.3782e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-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(2.9763e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-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(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<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(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-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(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-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(4.4327e-05, device='cuda:0', grad_fn=<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(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-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(5.0563e-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(4.4508e-05, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-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(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(2.6675e-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(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7275e-05, 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(4.2238e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9777e-05, device='cuda:0', grad_fn=<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(8.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4862e-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(0.0013, 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(0.0001, 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(0.0018, 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(4.2571e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, 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(5.1684e-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(0.0002, 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(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7891e-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(0.0007, 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(3.3669e-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(4.0542e-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(2.4707e-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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0463e-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(0.0001, 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(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-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(2.6372e-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(2.7160e-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(0.0009, 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(0.0005, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-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(9.8489e-05, device='cuda:0', grad_fn=<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(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5005e-05, 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(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3147e-05, device='cuda:0', grad_fn=<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(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(6.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1094e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-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(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8263e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1189e-05, device='cuda:0', grad_fn=<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(5.4518e-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(9.5667e-05, 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(0.0006, 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(8.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8033e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<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(0.0003, 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(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9726e-05, device='cuda:0', grad_fn=<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(0.0008, 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(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8185e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4218e-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(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2572e-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(4.6082e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-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(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-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(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-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(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0008, 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(8.2864e-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(4.2510e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5894e-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(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(6.8880e-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(2.4344e-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(0.0007, 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>)
			------------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
				 {'precision': {'support': 0.8796095444685467, 'agreement': 0.6507936507936508, 'direct_attack': 0.2426470588235294, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9700956937799043, 'agreement': 0.4659090909090909, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9226393629124006, 'agreement': 0.543046357615894, 'direct_attack': 0.26829268292682923, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.733044733044733, 'recall': 0.733044733044733, 'f1': 0.733044733044733, 'support': None}, 'macro_avg': {'precision': 0.5126934313113918, 'recall': 0.46362789687236294, 'f1': 0.4727202369040425, 'support': None}, 'weighted_avg': {'precision': 0.7114021758566964, 'recall': 0.733044733044733, 'f1': 0.7147361947294265, 'support': None}}
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(2.6706e-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(5.0987e-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(0.0009, 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(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5740e-05, device='cuda:0', grad_fn=<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(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-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(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5178e-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(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.7220e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-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(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(0.0009, 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(0.0005, 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(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, 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(9.3525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-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(9.9667e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<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(6.3612e-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(0.0028, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5934e-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(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3813e-05, device='cuda:0', grad_fn=<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(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4425e-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(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7513e-05, device='cuda:0', grad_fn=<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(0.0008, 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(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, 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(0.0003, 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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9939e-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(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-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(4.0149e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5357e-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(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8314e-05, device='cuda:0', grad_fn=<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(5.5650e-05, device='cuda:0', grad_fn=<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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3637e-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(8.9192e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-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(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3953e-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(0.0002, 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(4.7294e-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(6.3579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3823e-05, 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(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9695e-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(5.3832e-05, device='cuda:0', grad_fn=<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(6.9091e-05, device='cuda:0', grad_fn=<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(5.1321e-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(8.0079e-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(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, 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(0.0004, 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(2.5010e-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(7.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9224e-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(0.0011, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4714e-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(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6598e-05, device='cuda:0', grad_fn=<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(4.0572e-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(0.0002, 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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0009, 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(2.8038e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6805e-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(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(3.7423e-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(7.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-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(4.8686e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7540e-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(2.4859e-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(0.0004, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7252e-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(0.0015, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7395e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9677e-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(2.6675e-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(3.5183e-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(4.7869e-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(4.4841e-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(0.0002, device='cuda:0', grad_fn=<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(0.0003, 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(7.2300e-05, device='cuda:0', grad_fn=<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(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8365e-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(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1580e-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(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(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1454e-05, device='cuda:0', grad_fn=<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(7.2421e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4490e-05, device='cuda:0', grad_fn=<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(0.0002, 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(4.2207e-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(2.5585e-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(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, 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(8.4258e-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(2.3254e-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(8.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-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(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-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(4.0905e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-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(0.0010, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-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(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(2.6130e-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(7.2635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0005, 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(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(0.0006, 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(7.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4558e-05, device='cuda:0', grad_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(0.0003, 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(3.3306e-05, device='cuda:0', grad_fn=<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(3.1429e-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(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6866e-05, device='cuda:0', grad_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(0.0006, 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(0.0006, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6396e-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(7.3875e-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(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(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8262e-05, 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(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0163e-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(8.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6361e-05, 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(9.8430e-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(5.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5832e-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(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-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(4.0481e-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(9.4278e-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(0.0004, 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(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-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(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4254e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6228e-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(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-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(4.6945e-05, 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(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(2.5827e-05, device='cuda:0', grad_fn=<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(4.5871e-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(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9554e-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(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, 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(0.0006, 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(7.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0866e-05, device='cuda:0', grad_fn=<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(0.0002, 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(4.9473e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0070e-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(5.2653e-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(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0472e-05, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7897e-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(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0897e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-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(5.8435e-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(0.0004, 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(2.8462e-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(2.6857e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7770e-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(4.3449e-05, 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(0.0002, 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(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5417e-05, device='cuda:0', grad_fn=<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(0.0002, 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(5.1250e-05, 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(0.0005, 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(3.5637e-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(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3872e-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(9.9851e-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(7.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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
				 {'precision': {'support': 0.8789189189189189, 'agreement': 0.6666666666666666, 'direct_attack': 0.2426470588235294, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.972488038277512, 'agreement': 0.4659090909090909, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9233390119250426, 'agreement': 0.5484949832775918, 'direct_attack': 0.26829268292682923, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7344877344877345, 'recall': 0.7344877344877345, 'f1': 0.7344877344877345, 'support': None}, 'macro_avg': {'precision': 0.5157299093760694, 'recall': 0.46410636577188447, 'f1': 0.47394989183891045, 'support': None}, 'weighted_avg': {'precision': 0.7130012291280975, 'recall': 0.7344877344877345, 'f1': 0.7158500942179025, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2550e-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(7.5417e-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(7.0118e-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(4.5492e-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(3.9301e-05, device='cuda:0', grad_fn=<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(0.0003, 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(8.2805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0503e-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(0.0002, 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(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(0.0002, 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(1.9348e-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(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(7.1694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3188e-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(0.0005, 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(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1080e-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(5.9131e-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(0.0005, 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(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7465e-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(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(0.0002, 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(0.0001, 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(9.9060e-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(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5167e-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(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(4.1027e-05, 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(0.0001, 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(0.0006, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4149e-05, 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(0.0012, 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(0.0004, 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(0.0002, 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(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6086e-05, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6707e-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(0.0007, 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(7.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5289e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0917e-05, 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(5.3198e-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(7.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-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(0.0003, 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(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0500e-05, 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(2.5949e-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(2.7826e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5706e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-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(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, 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(3.3972e-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(0.0006, 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(2.4798e-05, device='cuda:0', grad_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(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, 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(2.6706e-05, 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(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(8.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-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(5.3379e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1172e-05, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1956e-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(2.5737e-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(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, 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(0.0002, 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(7.7460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6710e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8927e-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(0.0002, 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(0.0002, 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(0.0004, 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(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, 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(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0863e-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(0.0002, 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(5.7798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9804e-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(3.7666e-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(0.0003, 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(2.5161e-05, device='cuda:0', grad_fn=<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(5.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-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(4.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8380e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9162e-05, 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(6.9060e-05, device='cuda:0', grad_fn=<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(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4679e-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(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3728e-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(7.5971e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3760e-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(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5189e-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(2.4132e-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(4.4962e-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(0.0001, 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(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-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(0.0009, 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(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3983e-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(3.0157e-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(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(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5825e-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(6.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, 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(0.0004, 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(0.0007, 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(5.7629e-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(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5157e-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(0.0003, 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(0.0004, 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(3.7030e-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(4.0330e-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(2.8310e-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(6.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5224e-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(5.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9477e-05, 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(3.9543e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<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(2.5827e-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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<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(3.3730e-05, device='cuda:0', grad_fn=<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(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9534e-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(5.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8152e-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(9.5184e-05, device='cuda:0', grad_fn=<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(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(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2568e-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(0.0009, 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(2.7644e-05, device='cuda:0', grad_fn=<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(9.0036e-05, 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(7.1673e-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(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-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(0.0003, 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(4.8051e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-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(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-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(2.3042e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-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>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7022e-05, 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(6.5247e-05, device='cuda:0', grad_fn=<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(2.3799e-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(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0702e-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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4459e-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(6.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-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(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8266e-05, device='cuda:0', grad_fn=<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(2.4071e-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(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_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(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8931e-05, device='cuda:0', grad_fn=<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(3.2822e-05, 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(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-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(0.0006, 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(8.4854e-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(4.9349e-05, 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(0.0002, 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(0.0003, 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(0.0001, 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(0.0009, 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(0.0004, 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(7.0210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9940e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8851e-05, 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(0.0006, 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(6.5002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-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(3.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2118e-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(0.0003, 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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-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(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(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0107e-05, device='cuda:0', grad_fn=<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(0.0009, 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(0.0006, 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(0.0005, 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(7.7627e-05, device='cuda:0', grad_fn=<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(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6508e-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(0.0001, 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(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5476e-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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-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(8.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4488e-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(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, 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(8.7952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-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(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8797399783315276, 'agreement': 0.656, 'direct_attack': 0.2426470588235294, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9712918660287081, 'agreement': 0.4659090909090909, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9232518476407049, 'agreement': 0.5448504983388704, 'direct_attack': 0.26829268292682923, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7337662337662337, 'recall': 0.7337662337662337, 'f1': 0.7337662337662337, 'support': None}, 'macro_avg': {'precision': 0.5137607879252578, 'recall': 0.4638671313221237, 'f1': 0.47320356199429864, 'support': None}, 'weighted_avg': {'precision': 0.7121419739589832, 'recall': 0.7337662337662337, 'f1': 0.7153347271970357, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-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(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4429e-05, 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(4.2843e-05, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8301e-05, 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(0.0001, 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(2.8038e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5653e-05, device='cuda:0', grad_fn=<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(2.7553e-05, 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(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(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, 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(0.0005, 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(9.2702e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0120e-05, device='cuda:0', grad_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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-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(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0678e-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(1.6714e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-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(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(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(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0314e-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(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8442e-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(1.8682e-05, 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(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(5.8572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-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(0.0003, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5982e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-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(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-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(1.8712e-05, 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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<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(8.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-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(0.0003, 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(2.3799e-05, device='cuda:0', grad_fn=<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(0.0005, 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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5567e-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(2.2043e-05, device='cuda:0', grad_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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4904e-05, device='cuda:0', grad_fn=<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(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, device='cuda:0', grad_fn=<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(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7437e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9241e-05, 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(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, 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(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(0.0010, 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(2.2527e-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(0.0003, 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(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9563e-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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3937e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(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(0.0001, 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(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3357e-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(2.4283e-05, 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(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(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-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(7.6842e-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(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(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2461e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(4.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-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(0.0005, 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(0.0013, 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(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, 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(2.8340e-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(3.7181e-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(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(0.0006, 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(9.4699e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, 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(6.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1895e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9927e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2798e-05, device='cuda:0', grad_fn=<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(0.0002, 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(3.2337e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9989e-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(0.0012, 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(0.0005, 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(0.0006, 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(5.7981e-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(0.0007, 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(5.3319e-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(6.0735e-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(0.0002, 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(0.0008, 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(8.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9587e-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(6.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0930e-05, 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(0.0004, 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(2.4525e-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(8.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8707e-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(0.0003, 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(0.0001, 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(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<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(0.0002, 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(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(3.0399e-05, device='cuda:0', grad_fn=<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(6.1014e-05, 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(2.0468e-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(0.0001, 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(0.0004, 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(5.4287e-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(6.4671e-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(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, 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(0.0006, 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(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(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4857e-05, 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(0.0007, 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(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(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(0.0003, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, 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(3.7652e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1837e-05, 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(4.7990e-05, device='cuda:0', grad_fn=<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(5.9947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, 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(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(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(9.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3970e-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(0.0004, 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(5.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4982e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1820e-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(0.0002, 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(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(4.6446e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6163e-05, 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(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5519e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5261e-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(2.7250e-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(7.5630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0630e-05, device='cuda:0', grad_fn=<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(1.5351e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7020e-05, 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(2.7432e-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(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(6.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-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(5.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-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(0.0002, 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(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5523e-05, device='cuda:0', grad_fn=<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(2.3950e-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(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(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-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(2.0408e-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(3.9538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-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(5.3198e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2241e-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(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-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(0.0002, 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(5.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-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(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(8.8012e-05, 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(2.9097e-05, device='cuda:0', grad_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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-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(0.0003, 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(8.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9136e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-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
				 {'precision': {'support': 0.8797399783315276, 'agreement': 0.664, 'direct_attack': 0.2426470588235294, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9712918660287081, 'agreement': 0.4715909090909091, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9232518476407049, 'agreement': 0.5514950166112957, 'direct_attack': 0.26829268292682923, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7344877344877345, 'recall': 0.7344877344877345, 'f1': 0.7344877344877345, 'support': None}, 'macro_avg': {'precision': 0.5153607879252579, 'recall': 0.46500349495848736, 'f1': 0.47453246564878376, 'support': None}, 'weighted_avg': {'precision': 0.7131578469748562, 'recall': 0.7344877344877345, 'f1': 0.7161784755490896, 'support': None}}
Loss: tensor(1.9166e-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(2.1134e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-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(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2925e-05, 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(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(2.7947e-05, device='cuda:0', grad_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(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(0.0003, 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(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7526e-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(6.7275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1820e-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(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, 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(0.0006, 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(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6829e-05, 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(4.2207e-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(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, 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(7.2275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7577e-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(0.0004, 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(3.4460e-05, 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(3.0127e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-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(4.9655e-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(0.0002, 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(7.5396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, 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(7.6780e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-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(0.0003, 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(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4155e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6153e-05, device='cuda:0', grad_fn=<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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-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(0.0001, 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(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, 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(1.8712e-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(9.3371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<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(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0005, 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(2.7674e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-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(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7995e-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(9.7394e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3584e-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(2.2164e-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(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6079e-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(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5931e-05, 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(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, 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(4.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-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(0.0001, 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(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-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(7.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1561e-05, 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(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-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(0.0007, 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(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-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(8.1101e-05, 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(2.2981e-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(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8273e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4661e-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(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(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, 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(3.9392e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2240e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<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(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, 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(1.5684e-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(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2128e-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(1.6835e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, 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(5.0688e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9679e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3779e-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(7.2028e-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(2.5525e-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(2.1558e-05, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-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(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9391e-05, device='cuda:0', grad_fn=<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(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, 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(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5034e-05, 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(0.0003, 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(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7640e-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(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3378e-05, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, 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(0.0003, 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(0.0003, 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(8.9427e-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(9.3860e-05, 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(3.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3381e-05, device='cuda:0', grad_fn=<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(0.0004, 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(9.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4541e-05, 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(2.3375e-05, 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(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3841e-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(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-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(4.1117e-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(5.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-05, device='cuda:0', grad_fn=<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(0.0005, 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(6.9183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-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(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2187e-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(4.7476e-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(0.0007, 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(2.1467e-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(6.3447e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-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(0.0001, 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(1.7168e-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(0.0004, 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(4.3100e-05, device='cuda:0', grad_fn=<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(1.8076e-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(1.6320e-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(2.9219e-05, device='cuda:0', grad_fn=<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(0.0014, 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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7298e-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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(3.2700e-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(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-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(6.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4459e-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(8.0202e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, 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(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4820e-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(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(7.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2824e-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(2.5343e-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(0.0003, 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(0.0003, 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(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0491e-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(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7435e-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(5.1077e-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(0.0006, 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(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, device='cuda:0', grad_fn=<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(3.0944e-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(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6469e-05, device='cuda:0', grad_fn=<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(6.0009e-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(2.5676e-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(0.0003, 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(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, 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(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4180e-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(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-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(0.0003, 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(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-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(0.0002, 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(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6749e-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(6.6487e-05, device='cuda:0', grad_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(7.7498e-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(4.9796e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5789e-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(1.7562e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<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>)
			------------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
				 {'precision': {'support': 0.8786565547128927, 'agreement': 0.6587301587301587, 'direct_attack': 0.24444444444444444, 'undercutter_attack': 0.4678362573099415, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9700956937799043, 'agreement': 0.4715909090909091, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.10309278350515463}, 'f1': {'support': 0.9221148379761227, 'agreement': 0.5496688741721855, 'direct_attack': 0.2693877551020408, 'undercutter_attack': 0.47337278106508873, 'partial': 0.15624999999999997}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7337662337662337, 'recall': 0.7337662337662337, 'f1': 0.7337662337662337, 'support': None}, 'macro_avg': {'precision': 0.5144496120717456, 'recall': 0.4647642605087266, 'f1': 0.4741588496630875, 'support': None}, 'weighted_avg': {'precision': 0.7119778168230739, 'recall': 0.7337662337662337, 'f1': 0.7153476795825667, 'support': None}}
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.0892e-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(3.5279e-05, device='cuda:0', grad_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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-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(8.5824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6273e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4821e-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(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(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5557e-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(0.0001, 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(9.5444e-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(0.0001, 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(2.6312e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-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(5.4680e-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(5.4681e-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(8.4497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-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(4.7657e-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(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, 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(3.3003e-05, 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(6.1358e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(2.0680e-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(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(7.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-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(4.6960e-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(0.0004, 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(0.0001, 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(0.0003, 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(3.1035e-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(6.2704e-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(0.0002, device='cuda:0', grad_fn=<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(9.7336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9222e-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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1467e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2240e-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(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-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(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4279e-05, device='cuda:0', grad_fn=<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(3.2004e-05, device='cuda:0', grad_fn=<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(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(4.6597e-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(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, device='cuda:0', grad_fn=<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(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-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(1.7380e-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(2.9713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<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(3.2549e-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(0.0002, 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(6.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3328e-05, 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(2.8219e-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(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8956e-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(0.0001, 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(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3550e-05, device='cuda:0', grad_fn=<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(3.0218e-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(6.6449e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4874e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3786e-05, device='cuda:0', grad_fn=<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(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1772e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9979e-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(0.0004, 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(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2513e-05, 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(9.1461e-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(3.0376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7270e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6542e-05, device='cuda:0', grad_fn=<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(2.5464e-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(7.7049e-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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7141e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6953e-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(0.0015, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, 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(2.1982e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1909e-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(0.0002, 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(4.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-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(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0886e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0601e-05, device='cuda:0', grad_fn=<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(4.3872e-05, 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(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-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(1.4625e-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(2.6221e-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(0.0001, 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(2.6948e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-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(1.7622e-05, device='cuda:0', grad_fn=<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(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(3.0641e-05, 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(0.0006, 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(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, 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(6.4548e-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(0.0004, 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(7.7912e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5780e-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(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-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(0.0005, 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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7730e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8227e-05, 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(7.8241e-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(1.8894e-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(6.3869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.3396e-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(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2353e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3214e-05, device='cuda:0', grad_fn=<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(2.9128e-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(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9330e-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(2.4435e-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(2.4374e-05, 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(1.5563e-05, device='cuda:0', grad_fn=<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(4.3600e-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(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2690e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6890e-05, device='cuda:0', grad_fn=<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(1.6381e-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(0.0011, 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(3.1459e-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(2.0862e-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(6.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, 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(7.4569e-05, 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(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(6.7689e-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(3.6091e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0529e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-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(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-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(0.0003, 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(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8127e-05, 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(3.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7143e-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(0.0002, 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(4.8777e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, 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(9.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3776e-05, 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(0.0006, 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(9.8207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5973e-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(1.9257e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<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(2.4102e-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(9.9207e-05, device='cuda:0', grad_fn=<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(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1367e-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(9.0565e-05, device='cuda:0', grad_fn=<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(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-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(6.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2037e-05, device='cuda:0', grad_fn=<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(9.5358e-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(0.0001, 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(1.9863e-05, device='cuda:0', grad_fn=<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(0.0009, 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(7.3481e-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(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1824e-05, 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(1.4170e-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(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-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(2.7627e-05, device='cuda:0', grad_fn=<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(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5713e-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(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-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(4.2848e-05, device='cuda:0', grad_fn=<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(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8478e-05, 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(8.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4591e-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(0.0001, 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(8.5077e-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(0.0003, 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(1.6562e-05, 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>)
			------------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
				 {'precision': {'support': 0.8797399783315276, 'agreement': 0.6666666666666666, 'direct_attack': 0.2462686567164179, 'undercutter_attack': 0.4678362573099415, 'partial': 0.34375}, 'recall': {'support': 0.9712918660287081, 'agreement': 0.4772727272727273, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.1134020618556701}, 'f1': {'support': 0.9232518476407049, 'agreement': 0.5562913907284768, 'direct_attack': 0.27049180327868855, 'undercutter_attack': 0.47337278106508873, 'partial': 0.17054263565891473}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7359307359307359, 'recall': 0.7359307359307359, 'f1': 0.7359307359307359, 'support': None}, 'macro_avg': {'precision': 0.5208523118049107, 'recall': 0.46820171426495405, 'f1': 0.4787900916743747, 'support': None}, 'weighted_avg': {'precision': 0.715265449082292, 'recall': 0.7359307359307359, 'f1': 0.717962350543493, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.2025e-05, device='cuda:0', grad_fn=<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(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, 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(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3895e-05, 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(0.0003, 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(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5997e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7374e-05, 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(1.9802e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6103e-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(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0002, 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(2.0832e-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(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4204e-05, 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(5.8305e-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(0.0002, 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(0.0001, 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(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7026e-05, device='cuda:0', grad_fn=<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(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-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(9.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3287e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1068e-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(1.8288e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(4.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, 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(0.0008, 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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8790e-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(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5451e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1458e-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(7.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-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(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4512e-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(0.0003, 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(0.0003, 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(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2787e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8844e-05, 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(8.2078e-05, 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(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8337e-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(4.5198e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6464e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, 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(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9099e-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(0.0003, 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(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6839e-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(3.5662e-05, device='cuda:0', grad_fn=<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(6.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-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(5.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(9.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, 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(0.0003, 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(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5320e-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(5.2864e-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(2.7099e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8444e-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(0.0002, 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(7.8266e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0059e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-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(0.0001, 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(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-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(3.3475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-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(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(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8211e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_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(2.1195e-05, 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(1.3202e-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(0.0003, 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(2.3920e-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(8.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7911e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5434e-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(0.0005, 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(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, 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(2.0589e-05, 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(0.0003, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<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(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-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(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3650e-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(9.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-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(9.3000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-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(4.2479e-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(9.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4710e-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(0.0004, 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(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.5315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6149e-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(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-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(0.0002, 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(0.0002, 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(5.4233e-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(9.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8893e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-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(0.0005, 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(0.0004, 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(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-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(0.0004, 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(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(4.4387e-05, 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(6.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8942e-05, 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(1.5896e-05, device='cuda:0', grad_fn=<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(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6599e-05, device='cuda:0', grad_fn=<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(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0390e-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(2.6887e-05, device='cuda:0', grad_fn=<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(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(9.9178e-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(1.5927e-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(4.4114e-05, 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(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(2.0832e-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(6.1091e-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(2.2043e-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(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(9.7632e-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(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8059e-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(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-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(5.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0854e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7466e-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(0.0004, 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(1.4867e-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(0.0002, 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(4.3329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8203e-05, 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(0.0004, 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(0.0001, 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(0.0001, 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(2.3738e-05, 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(0.0001, 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(9.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0943e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8756e-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(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9559e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2948e-05, 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(3.5062e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9613e-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(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, 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(2.0044e-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(4.0179e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<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(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0808e-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(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8766e-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(4.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-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
				 {'precision': {'support': 0.8797399783315276, 'agreement': 0.6666666666666666, 'direct_attack': 0.2462686567164179, 'undercutter_attack': 0.4678362573099415, 'partial': 0.34375}, 'recall': {'support': 0.9712918660287081, 'agreement': 0.4772727272727273, 'direct_attack': 0.3, 'undercutter_attack': 0.47904191616766467, 'partial': 0.1134020618556701}, 'f1': {'support': 0.9232518476407049, 'agreement': 0.5562913907284768, 'direct_attack': 0.27049180327868855, 'undercutter_attack': 0.47337278106508873, 'partial': 0.17054263565891473}, 'support': {'support': 836, 'agreement': 176, 'direct_attack': 110, 'undercutter_attack': 167, 'partial': 97}, 'micro_avg': {'precision': 0.7359307359307359, 'recall': 0.7359307359307359, 'f1': 0.7359307359307359, 'support': None}, 'macro_avg': {'precision': 0.5208523118049107, 'recall': 0.46820171426495405, 'f1': 0.4787900916743747, 'support': None}, 'weighted_avg': {'precision': 0.715265449082292, 'recall': 0.7359307359307359, 'f1': 0.717962350543493, '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
				 {'precision': {'support': 0.0, 'agreement': 0.17073170731707318, 'direct_attack': 0.16666666666666666, 'undercutter_attack': 0.1134453781512605, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 0.034482758620689655, 'direct_attack': 0.007633587786259542, 'undercutter_attack': 0.9938650306748467, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.05737704918032787, 'direct_attack': 0.0145985401459854, 'undercutter_attack': 0.20364550597108735, 'partial': 0.0}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 163, 'partial': 116}, 'micro_avg': {'precision': 0.11486486486486487, 'recall': 0.11486486486486487, 'f1': 0.11486486486486487, 'support': None}, 'macro_avg': {'precision': 0.09016875042700007, 'recall': 0.20719627541635915, 'f1': 0.05512421905948013, 'support': None}, 'weighted_avg': {'precision': 0.05066450443064504, 'recall': 0.11486486486486487, 'f1': 0.03159065352433641, 'support': None}}
Loss: tensor(0.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0724, 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
				 {'precision': {'support': 0.6723044397463002, 'agreement': 0.39649122807017545, 'direct_attack': 0.225, 'undercutter_attack': 0.11578947368421053, 'partial': 0.10526315789473684}, 'recall': {'support': 0.7335640138408305, 'agreement': 0.5566502463054187, 'direct_attack': 0.06923076923076923, 'undercutter_attack': 0.13414634146341464, 'partial': 0.017241379310344827}, 'f1': {'support': 0.7015995587424159, 'agreement': 0.46311475409836067, 'direct_attack': 0.10588235294117648, 'undercutter_attack': 0.12429378531073446, 'partial': 0.029629629629629627}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 130, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.5283783783783784, 'recall': 0.5283783783783784, 'f1': 0.5283783783783784, 'support': None}, 'macro_avg': {'precision': 0.3029696598790846, 'recall': 0.3021665500301555, 'f1': 0.2849040161444634, 'support': None}, 'weighted_avg': {'precision': 0.4890713976745188, 'recall': 0.5283783783783784, 'f1': 0.4999223217716164, 'support': None}}
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6542, 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
				 {'precision': {'support': 0.9080459770114943, 'agreement': 0.4470198675496689, 'direct_attack': 0.04878048780487805, 'undercutter_attack': 0.20531757754800592, 'partial': 0.16071428571428573}, 'recall': {'support': 0.3648960739030023, 'agreement': 0.6650246305418719, 'direct_attack': 0.015267175572519083, 'undercutter_attack': 0.8475609756097561, 'partial': 0.15517241379310345}, 'f1': {'support': 0.5205930807248764, 'agreement': 0.5346534653465347, 'direct_attack': 0.023255813953488372, 'undercutter_attack': 0.33055885850178357, 'partial': 0.15789473684210528}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.41216216216216217, 'recall': 0.41216216216216217, 'f1': 0.41216216216216217, 'support': None}, 'macro_avg': {'precision': 0.3539756391256666, 'recall': 0.40958425388405056, 'f1': 0.3133911910737577, 'support': None}, 'weighted_avg': {'precision': 0.6323094817349364, 'recall': 0.41216216216216217, 'f1': 0.4290150103168738, 'support': None}}
Loss: tensor(1.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, 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(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, 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
				 {'precision': {'support': 0.890625, 'agreement': 0.47619047619047616, 'direct_attack': 0.23015873015873015, 'undercutter_attack': 0.4020979020979021, 'partial': 0.4}, 'recall': {'support': 0.8556581986143187, 'agreement': 0.541871921182266, 'direct_attack': 0.22137404580152673, 'undercutter_attack': 0.7012195121951219, 'partial': 0.017241379310344827}, 'f1': {'support': 0.872791519434629, 'agreement': 0.5069124423963133, 'direct_attack': 0.22568093385214008, 'undercutter_attack': 0.5111111111111111, 'partial': 0.03305785123966942}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.6736486486486486, 'recall': 0.6736486486486486, 'f1': 0.6736486486486486, 'support': None}, 'macro_avg': {'precision': 0.4798144216894217, 'recall': 0.4674730114207157, 'f1': 0.42991077160677255, 'support': None}, 'weighted_avg': {'precision': 0.6827315988253487, 'recall': 0.6736486486486486, 'f1': 0.6594336600929017, 'support': None}}
Loss: tensor(0.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, 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(2.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, 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(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4320, 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(0.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9880, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7940, 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(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0098, 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(1.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5212, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, 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(0.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, 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(1.8317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4105, 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(0.0151, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, 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(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3450, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7925, 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(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, 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(1.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, 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(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, 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(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8970, 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(0.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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
				 {'precision': {'support': 0.9021065675340768, 'agreement': 0.484375, 'direct_attack': 0.2669172932330827, 'undercutter_attack': 0.3472222222222222, 'partial': 0.22784810126582278}, 'recall': {'support': 0.839677047289504, 'agreement': 0.6108374384236454, 'direct_attack': 0.5419847328244275, 'undercutter_attack': 0.15337423312883436, 'partial': 0.15517241379310345}, 'f1': {'support': 0.8697729988052568, 'agreement': 0.5403050108932462, 'direct_attack': 0.3576826196473552, 'undercutter_attack': 0.2127659574468085, 'partial': 0.18461538461538463}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 163, 'partial': 116}, 'micro_avg': {'precision': 0.6527027027027027, 'recall': 0.6527027027027027, 'f1': 0.6527027027027027, 'support': None}, 'macro_avg': {'precision': 0.4456938368510409, 'recall': 0.460209173091903, 'f1': 0.43302839428161033, 'support': None}, 'weighted_avg': {'precision': 0.6746272205639433, 'recall': 0.6527027027027027, 'f1': 0.6531944365057464, 'support': None}}
Loss: tensor(0.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3305, 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(0.8420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, 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(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3475, 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(0.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, 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(0.5209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2180, 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(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4773, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0040, 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(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, 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(0.0802, 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(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4405, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, 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(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, 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(0.9765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5909, 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(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, 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(0.1723, 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(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, 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(0.6421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2996, 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(0.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3109, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, 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(2.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, 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(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, 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(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, 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(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4996, 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(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, 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(5.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3913, 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(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, 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(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, 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(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6039, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, 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(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, 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(0.5514, 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(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.9814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.5711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, 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(0.3306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, 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(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9437, 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(0.2379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, 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(4.8519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, 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(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, 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(2.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, 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(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, 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(0.6022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(2.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, 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
				 {'precision': {'support': 0.8872266973532796, 'agreement': 0.5140562248995983, 'direct_attack': 0.2558139534883721, 'undercutter_attack': 0.43006993006993005, 'partial': 0.12121212121212122}, 'recall': {'support': 0.889273356401384, 'agreement': 0.6336633663366337, 'direct_attack': 0.08396946564885496, 'undercutter_attack': 0.75, 'partial': 0.034482758620689655}, 'f1': {'support': 0.8882488479262671, 'agreement': 0.5676274944567629, 'direct_attack': 0.1264367816091954, 'undercutter_attack': 0.5466666666666667, 'partial': 0.053691275167785234}, 'support': {'support': 867, 'agreement': 202, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7006756756756757, 'recall': 0.7006756756756757, 'f1': 0.7006756756756757, 'support': None}, 'macro_avg': {'precision': 0.4416757854046603, 'recall': 0.47827778940151244, 'f1': 0.4365342131653354, 'support': None}, 'weighted_avg': {'precision': 0.6697085179284212, 'recall': 0.7006756756756757, 'f1': 0.6737954355918518, 'support': None}}
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, 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(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3480, 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(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.2855, 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(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, 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(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3451, 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(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, 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(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, 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(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, 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(0.0363, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, 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(1.9793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, 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(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, 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(0.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, 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(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, 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(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, 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(0.1044, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, 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(0.7286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(1.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.4073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, 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(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1271, 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(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, 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(0.0355, 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(0.0052, 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(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.0002, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, 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(0.0116, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, 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(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6805, device='cuda:0', grad_fn=<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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, 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(0.0191, 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(0.0005, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<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(0.0064, device='cuda:0', grad_fn=<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(0.0085, device='cuda:0', grad_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(0.0008, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, 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(0.0235, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4501, 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(0.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, 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(0.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0007, 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(0.4261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4443, 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(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, 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(0.0192, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, 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(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, 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(0.0611, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, 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(0.0028, 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(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(5.2940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, 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(0.7253, 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(0.0005, 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(0.0057, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, 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(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2849, 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(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, 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(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0665, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, 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(0.1667, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6314, 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(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, 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(1.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, 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(0.0028, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, 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(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, 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(0.0192, 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(0.0030, 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(0.0007, 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(0.0029, 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(0.0013, 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(0.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, 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(2.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7766, 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(1.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, 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(0.0191, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2691, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, 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(0.0500, 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(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7520, 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(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, 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(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, 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(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, 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(0.0087, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, 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(0.0011, 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
				 {'precision': {'support': 0.8833151581243184, 'agreement': 0.6067415730337079, 'direct_attack': 0.25252525252525254, 'undercutter_attack': 0.41901408450704225, 'partial': 0.5}, 'recall': {'support': 0.9353348729792148, 'agreement': 0.5320197044334976, 'direct_attack': 0.19083969465648856, 'undercutter_attack': 0.725609756097561, 'partial': 0.008620689655172414}, 'f1': {'support': 0.9085810431856421, 'agreement': 0.5669291338582678, 'direct_attack': 0.21739130434782608, 'undercutter_attack': 0.53125, 'partial': 0.01694915254237288}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7182432432432433, 'recall': 0.7182432432432433, 'f1': 0.7182432432432433, 'support': None}, 'macro_avg': {'precision': 0.5323192136380642, 'recall': 0.47848494356438687, 'f1': 0.4482201267868218, 'support': None}, 'weighted_avg': {'precision': 0.7080530974334226, 'recall': 0.7182432432432433, 'f1': 0.6888426757678885, 'support': None}}
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, 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(0.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, 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(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, 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(0.2299, 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(2.2220, 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(0.0011, 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(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0866, 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(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2483, 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(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, 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(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, 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(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, 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(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, 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(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, 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(0.0013, 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(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3809, 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(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(0.0021, 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(0.0573, 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(0.0006, 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(0.0030, 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(0.0075, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, 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(0.0118, 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(0.0052, 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(0.0005, 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(0.0005, 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(0.0013, 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(0.0013, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2304, 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(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, 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(0.0221, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, 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(0.0027, 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(0.1357, 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(0.0032, 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(0.0792, 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(0.0061, 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(0.0006, 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(0.2360, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2707, 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(1.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, 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(0.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3989, 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(0.3258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, 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(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, 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(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1198, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5516, 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(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4373, 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(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, 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(0.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3174, 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(1.7732, 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(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, 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(0.0181, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4318, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, 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(0.0033, 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(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, 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(0.0037, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9545, 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(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, 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(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, 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(2.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4929, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, 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(2.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, 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(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4220, 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(0.0182, 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(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, 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(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, 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(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, 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(0.0464, 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(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, 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(0.7414, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, 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(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0059, 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(0.2285, 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
				 {'precision': {'support': 0.9080459770114943, 'agreement': 0.5364583333333334, 'direct_attack': 0.25396825396825395, 'undercutter_attack': 0.41295546558704455, 'partial': 0.14814814814814814}, 'recall': {'support': 0.9122401847575058, 'agreement': 0.5073891625615764, 'direct_attack': 0.12213740458015267, 'undercutter_attack': 0.6219512195121951, 'partial': 0.13793103448275862}, 'f1': {'support': 0.9101382488479263, 'agreement': 0.5215189873417722, 'direct_attack': 0.16494845360824742, 'undercutter_attack': 0.49635036496350365, 'partial': 0.14285714285714285}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.6939189189189189, 'recall': 0.6939189189189189, 'f1': 0.6939189189189189, 'support': None}, 'macro_avg': {'precision': 0.4519152356096548, 'recall': 0.4603298011788377, 'f1': 0.4471626395237185, 'support': None}, 'weighted_avg': {'precision': 0.6847625544391368, 'recall': 0.6939189189189189, 'f1': 0.684884603905951, 'support': None}}
Loss: tensor(0.0020, 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(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0003, 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(0.0044, 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(0.0008, 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(0.0043, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4616, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, 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(0.0004, 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(0.8031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2654, device='cuda:0', grad_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(0.0469, 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(0.0008, 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(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0099, 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(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, 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(0.0311, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, 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(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, 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(0.0143, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, 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(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0059, 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(0.0045, 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(0.0082, 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(0.6511, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, 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(3.0404, 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(0.0015, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, 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(0.0536, 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(0.0003, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, 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(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, 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(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, 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(0.0135, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, 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(0.0140, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7024, 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(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, 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(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, 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(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0016, 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(0.0051, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, 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(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, 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(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6185, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, 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(1.7485, 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(0.0204, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, 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(0.0006, 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(0.0004, 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(0.0125, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, 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(0.0048, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, 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(0.4691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, 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(0.0044, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, 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(0.0043, 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(0.0015, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0055, 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(0.0036, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0329, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, 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(0.0111, 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(0.0442, 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(1.0695, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(1.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, 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(0.3745, 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(0.0011, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0060, 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(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, 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(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8921, 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(0.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3519, 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(0.3848, 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(0.0020, 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(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, 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(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, 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(0.0196, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, 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(0.0779, 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(0.0007, 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(0.0013, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, 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(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, 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(0.0007, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2054, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, 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(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.1962, 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(0.0592, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0232, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, 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(0.5147, 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(0.0039, 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(0.0229, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, 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(0.0057, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0851, 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(0.0151, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<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(0.1597, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, 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(0.0087, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, 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(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1244, 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(0.0716, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, 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(0.3492, 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(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, 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(0.0035, 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(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, 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(0.0033, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, 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(0.0118, 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(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0020, 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(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0011, 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(0.0004, 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(0.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5933, 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(0.0004, 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(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, 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(0.0032, 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(0.0013, 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(0.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0018, 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(0.0014, 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(0.0023, 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(0.0344, 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(0.0039, 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(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(0.0042, 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
				 {'precision': {'support': 0.8825466520307355, 'agreement': 0.5585106382978723, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.4066666666666667, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9273356401384083, 'agreement': 0.5172413793103449, 'direct_attack': 0.13846153846153847, 'undercutter_attack': 0.7439024390243902, 'partial': 0.04310344827586207}, 'f1': {'support': 0.9043869516310461, 'agreement': 0.5370843989769821, 'direct_attack': 0.18652849740932642, 'undercutter_attack': 0.5258620689655172, 'partial': 0.0746268656716418}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 130, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7121621621621622, 'recall': 0.7121621621621622, 'f1': 0.7121621621621621, 'support': None}, 'macro_avg': {'precision': 0.4822432040974675, 'recall': 0.4740088890421088, 'f1': 0.4456977565309027, 'support': None}, 'weighted_avg': {'precision': 0.685543256475357, 'recall': 0.7121621621621622, 'f1': 0.6839719057080488, 'support': None}}
Loss: tensor(0.0128, 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(0.0806, 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(0.0018, 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(0.0002, 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(0.0028, 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(0.0202, 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(0.2246, device='cuda:0', grad_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(0.0085, 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(0.0014, 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(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, 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(0.0015, 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(0.0016, 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(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, 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(0.0007, 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(0.0046, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0335, 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(0.0013, 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(0.0079, 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(0.0009, 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(0.0012, 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(0.0028, 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(0.0369, 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(0.0070, 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(0.0027, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0813, 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(0.0146, 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(0.0015, 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(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, 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(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0005, 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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0004, 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(0.0005, 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(0.4079, 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(0.0005, 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(0.0030, 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(0.0005, 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(0.0005, 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(0.0030, 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(0.0027, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, 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(0.0003, 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(0.0121, 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(0.0057, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, 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(0.0001, 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(0.0003, 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(0.0055, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, 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(0.0008, 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(0.0003, 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(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0025, 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(0.0028, 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(0.0075, 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(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0003, 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(0.0011, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, 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(0.0438, 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(0.0015, 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(0.0026, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, 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(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0056, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, 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(0.0108, 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(0.0023, 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(0.0004, 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(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.0014, 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(0.0004, 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(0.0040, 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(0.0014, 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(0.0002, 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(0.0002, 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(0.0145, 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(0.0185, 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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, 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(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, 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(0.0081, 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(0.0002, 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(0.0022, 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(0.0004, 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(0.0006, 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(0.0109, 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(0.0010, 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(0.0173, 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(0.0069, 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(0.0011, 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(0.0009, 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(0.0005, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0026, 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(0.0013, 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(8.5138e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0023, 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(0.0191, device='cuda:0', grad_fn=<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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, 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(0.0017, 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(0.0002, 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(0.0027, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2304, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, 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(0.0020, 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(0.0041, 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(0.0003, 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(0.0017, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5168e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0002, 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(0.0140, 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(0.0060, 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(0.0115, device='cuda:0', grad_fn=<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(0.0063, 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(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(0.0021, 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(0.0007, 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(0.0002, 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(0.0063, 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(0.0087, 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(0.0028, 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(0.0004, 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(0.0126, 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(0.0060, 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(0.0027, 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(0.0002, 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(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0036, 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(0.4844, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, 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(0.0008, 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(0.0037, 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(0.0003, 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(0.0172, 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(0.0466, 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(0.0076, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0051, 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(0.0005, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0024, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, 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(0.0018, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0001, 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(0.0043, 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(0.0006, 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(0.0001, 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(0.0011, 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(0.0004, 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(0.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6015e-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(0.0116, 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(0.0015, 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(0.0016, 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(0.0033, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, 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(0.0048, 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(0.0099, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(0.0009, 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(0.0004, 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(0.0005, 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(0.0018, 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(0.0001, 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(0.0005, 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(0.0001, 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(0.0020, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0067, 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(0.0064, 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(0.0012, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, 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(0.0044, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(8.7952e-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(0.0001, 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(9.8246e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5017e-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(0.0036, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, 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(0.0012, 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(0.0011, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0004, 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(0.0004, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0047, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0070, 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(0.0026, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-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(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, 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(0.0192, 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(0.0011, 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(0.0022, 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(0.0150, 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(0.0135, 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(0.0001, 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(0.0241, 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(0.0122, 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(0.0070, 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(0.0001, 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(0.0531, device='cuda:0', grad_fn=<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(0.6173, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8708288482238966, 'agreement': 0.580110497237569, 'direct_attack': 0.2440944881889764, 'undercutter_attack': 0.39461883408071746, 'partial': 0.15}, 'recall': {'support': 0.9341801385681293, 'agreement': 0.5172413793103449, 'direct_attack': 0.2366412213740458, 'undercutter_attack': 0.5365853658536586, 'partial': 0.02586206896551724}, 'f1': {'support': 0.9013927576601671, 'agreement': 0.546875, 'direct_attack': 0.24031007751937986, 'undercutter_attack': 0.4547803617571059, 'partial': 0.04411764705882353}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7, 'recall': 0.7, 'f1': 0.7, 'support': None}, 'macro_avg': {'precision': 0.44793053354623186, 'recall': 0.4501020348143392, 'f1': 0.4374951687990952, 'support': None}, 'weighted_avg': {'precision': 0.6662122163804828, 'recall': 0.7, 'f1': 0.6775702700511705, 'support': None}}
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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4583e-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(0.0008, 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(0.0079, 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(0.0055, device='cuda:0', grad_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(0.0001, 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(0.0002, 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(0.0038, 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(0.0039, 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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, 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(0.0017, 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(0.0001, 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(0.0012, 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(0.0790, 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(0.0002, 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(0.0121, 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(0.0030, 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(0.0030, 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(0.0032, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0002, 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(0.0003, 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(0.0123, 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(0.0015, 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(0.0036, 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(0.0013, 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(0.0004, 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(0.1039, 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(0.0069, 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(0.0038, 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(0.0006, 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(0.0050, 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(0.0015, 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(0.0011, 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(0.0013, 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(0.0002, 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(0.0083, device='cuda:0', grad_fn=<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(0.0001, 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(0.0017, 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(0.0003, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0004, 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(0.0037, 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(0.0008, 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(0.0102, 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(0.0054, 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(0.0113, 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(0.0023, 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(0.0002, 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(0.0019, 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(0.0053, 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(0.0003, 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(0.0049, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9669e-05, 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(0.0012, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(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(0.0001, 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(0.0054, 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(0.0009, 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(0.0002, 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(0.0020, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.0005, 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(0.0022, 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(0.0028, 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(0.0018, 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(0.0022, 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(0.0022, 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(0.0013, 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(0.0056, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3048e-05, 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(0.0019, 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(0.0012, 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(0.0006, 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(0.0015, 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(0.0038, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0034, 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(0.0025, 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(0.0004, 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(0.0002, 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(0.0009, 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(0.0035, 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(0.0020, 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(0.0019, 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(0.0024, 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(0.0003, 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(0.0015, 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(0.0060, 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(0.0027, 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(0.0002, 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(0.0039, 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(0.0106, 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(0.0015, 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(0.0002, 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(0.0003, 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(0.0004, 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(0.0013, 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(0.0043, 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(0.0079, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-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(0.0023, 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(0.0001, 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(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0123, 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(0.0022, 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(0.0020, 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(0.0004, 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(0.0004, 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(0.0004, 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(0.0006, 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(0.0022, 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(0.0003, 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(0.0014, 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(0.0032, 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(0.0011, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5500e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0025, 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(0.0002, 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(0.0003, 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(0.0018, 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(0.0002, 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(0.0014, 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(0.0077, 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(0.0004, 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(8.6046e-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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0005, 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(0.0002, 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(8.7681e-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(0.0005, 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(0.0013, 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(0.0019, 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(0.0005, 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(0.0415, 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(0.0004, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, 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(0.0010, 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(0.0011, 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(0.0020, 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(0.0023, 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(0.0119, 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(0.0001, 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(0.0001, 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(0.0057, 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(0.0056, 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(0.0002, 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(0.0004, 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(0.0010, 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(0.0015, 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(0.0009, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0163e-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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0025, 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(0.0011, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.0003, 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(0.0043, 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(0.0012, 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(0.0035, 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(0.0026, 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(0.0003, 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(0.0033, 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(7.9749e-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(0.0002, 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(0.0001, 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(0.0008, 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(0.0026, 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(0.0014, 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(0.0001, 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(0.0014, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1535e-05, 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(0.0043, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3442e-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(0.0008, 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(0.0001, 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(0.0071, device='cuda:0', grad_fn=<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(0.0018, 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(8.5864e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0012, 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(0.0269, 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(0.0062, 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(0.0056, 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(0.0006, 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(0.0003, 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(0.0062, device='cuda:0', grad_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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4220e-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(0.0001, 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(0.0036, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5378e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0006, 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(0.0007, 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(0.0029, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, 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(0.0048, 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(9.2404e-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(0.0002, 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(0.0020, 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(0.0002, 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(0.0034, 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(0.0013, 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(6.3764e-05, 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(0.0012, 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(0.0008, 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(0.0013, 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(0.0019, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6490e-05, device='cuda:0', grad_fn=<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(0.0028, 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(0.0001, 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(0.0013, 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(0.0047, device='cuda:0', grad_fn=<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(0.0001, 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(0.0005, 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(0.0006, 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(0.0021, 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(0.0182, 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(0.0039, 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(0.0007, 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(0.0062, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2888e-05, 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(0.0002, 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(0.0006, 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(0.0004, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0007, 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(0.0048, 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(0.0039, 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>)
			------------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
				 {'precision': {'support': 0.8807439824945296, 'agreement': 0.5674157303370787, 'direct_attack': 0.23893805309734514, 'undercutter_attack': 0.4065040650406504, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9284890426758939, 'agreement': 0.4975369458128079, 'direct_attack': 0.20610687022900764, 'undercutter_attack': 0.6097560975609756, 'partial': 0.06086956521739131}, 'f1': {'support': 0.9039865244244808, 'agreement': 0.5301837270341208, 'direct_attack': 0.22131147540983606, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09722222222222222}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 115}, 'micro_avg': {'precision': 0.7027027027027027, 'recall': 0.7027027027027027, 'f1': 0.7027027027027027, 'support': None}, 'macro_avg': {'precision': 0.4669962282628862, 'recall': 0.46055170429921527, 'f1': 0.44810176542788804, 'support': None}, 'weighted_avg': {'precision': 0.678727431346796, 'recall': 0.7027027027027027, 'f1': 0.6834837649312131, 'support': None}}
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(0.0004, 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(0.0003, 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(0.0016, 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(8.7378e-05, 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(0.0082, 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(0.0067, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7821e-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(0.0013, 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(0.0108, 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(0.0014, 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(0.0007, 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(0.0006, 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(8.0203e-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(0.0011, 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(0.0010, 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(0.0041, 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(0.0001, 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(0.0018, 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(0.0015, 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(0.0019, 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(0.0358, 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(0.0008, 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(0.0045, 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(0.0034, 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(0.0001, 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(0.0021, 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(0.0034, 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(0.0018, 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(0.0017, 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(0.0011, 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(0.0012, 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(0.0007, 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(0.0003, 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(7.5238e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0023, 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(0.0014, 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(0.0011, 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(0.0001, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-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(0.0028, 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(0.0003, 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(0.0002, 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(0.0001, 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(0.0018, 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(0.0039, 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(0.0004, 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(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(0.0018, 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(0.0004, 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(0.0056, 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(0.0012, 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(0.0005, 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(0.0025, 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(0.0035, 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(0.0002, 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(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(9.4280e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0355e-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(0.0002, 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(0.0017, 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(0.0011, 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(0.0013, 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(0.0017, 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(0.0023, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(9.5551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4492e-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(0.0011, 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(0.0011, 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(0.0036, 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(0.0001, 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(0.0009, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-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(0.0045, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6460e-05, 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(7.3119e-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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5752e-05, 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(0.0010, 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(9.9730e-05, 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(0.0010, 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(0.0001, 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(0.0012, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-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(0.0016, 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(0.0015, 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(0.0010, 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(0.0003, 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(0.0001, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-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(0.0005, 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(0.0031, 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(0.0029, 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(0.0001, 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(0.0119, 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(8.6076e-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(0.0017, 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(0.0037, 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(0.0006, 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(0.0065, 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(0.0010, 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(0.0001, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, 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(0.0001, 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(0.0006, 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(7.8839e-05, 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(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(0.0010, 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(0.0014, 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(0.0020, 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(6.1917e-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(0.0026, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8943e-05, device='cuda:0', grad_fn=<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(0.0016, 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(0.0023, 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(6.8426e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3160e-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(0.0019, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, 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(0.0049, 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(0.0009, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5247e-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(0.0006, 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(0.0007, 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(0.0009, 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(0.0002, 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(0.0003, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9031e-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(0.0002, 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(0.0029, 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(0.0002, 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(7.6176e-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(0.0018, 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(0.0014, 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(0.0007, 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(0.0007, 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(0.0019, 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(0.0020, 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(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(0.0033, 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(9.6793e-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(0.0097, 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(0.0005, 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(0.0018, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2888e-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(0.0002, 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(8.8437e-05, 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(9.0254e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3250e-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(0.0003, 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(0.0012, 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(0.0015, 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(0.0015, 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(9.6096e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-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(0.0027, 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(0.0002, 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(0.0097, 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(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(0.0067, 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(0.0048, 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(0.0027, 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(0.0162, 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(0.0002, 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(0.0038, device='cuda:0', grad_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(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(0.0011, 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(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(0.0012, 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(0.0026, 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(9.0920e-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(0.0002, 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(6.2643e-05, device='cuda:0', grad_fn=<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(0.0064, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9275e-05, 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(9.1828e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2191e-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(9.5370e-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(0.0059, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, 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(0.0013, 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(8.8619e-05, 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(0.0001, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1162e-05, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5490e-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(0.0011, 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(0.0006, 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(0.0002, 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(0.0025, 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(0.0011, 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(0.0006, 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(0.0002, 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(0.0017, 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(0.0023, 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(8.9467e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6742e-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(0.0003, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9548e-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(0.0031, 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(6.9970e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8821e-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(9.0253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4663e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0011, 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(0.0031, 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(0.0012, 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(0.0007, 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(0.0018, 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(0.0027, 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(0.0028, 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(0.0009, 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(0.0011, 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(0.0006, 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(7.4118e-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(8.3079e-05, 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(0.0011, 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(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(7.1514e-05, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2019e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0006, 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(0.0081, device='cuda:0', grad_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(0.0019, 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(0.0035, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8807439824945296, 'agreement': 0.5722222222222222, 'direct_attack': 0.23214285714285715, 'undercutter_attack': 0.40408163265306124, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9284890426758939, 'agreement': 0.5073891625615764, 'direct_attack': 0.2, 'undercutter_attack': 0.6036585365853658, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9039865244244808, 'agreement': 0.5378590078328982, 'direct_attack': 0.21487603305785125, 'undercutter_attack': 0.4841075794621027, 'partial': 0.09655172413793105}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 130, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7027027027027027, 'recall': 0.7027027027027027, 'f1': 0.7027027027027027, 'support': None}, 'macro_avg': {'precision': 0.4661140009714996, 'recall': 0.4599763138818086, 'f1': 0.4474761737830528, 'support': None}, 'weighted_avg': {'precision': 0.6785230426469877, 'recall': 0.7027027027027027, 'f1': 0.6834251504023032, 'support': None}}
Loss: tensor(6.2432e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0101e-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(0.0043, 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(0.0017, 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(5.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-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(0.0020, 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(0.0004, 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(0.0043, device='cuda:0', grad_fn=<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(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6954e-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(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(0.0004, 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(0.0007, 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(0.0019, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0008, 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(0.0017, 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(0.0021, 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(0.0008, 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(0.0005, 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(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(0.0011, 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(0.0064, 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(0.0002, 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(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(0.0047, 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(0.0002, 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(0.0005, 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(0.0013, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0014, device='cuda:0', grad_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(0.0036, 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(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(0.0003, 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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0557e-05, 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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-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(0.0018, 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(0.0024, 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(0.0011, 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(0.0014, 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(0.0001, 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(0.0001, 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(8.3564e-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(0.0012, 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(0.0008, 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(0.0024, 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(0.0016, 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(0.0017, 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(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(5.3712e-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(0.0014, 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(0.0016, 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(0.0010, 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(0.0018, 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(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(5.8405e-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(0.0007, 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(0.0026, 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(6.9789e-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(0.0015, 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(0.0002, 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(0.0009, 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(0.0006, 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(0.0001, 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(0.0005, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4007e-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(9.2706e-05, 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(0.0029, 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(0.0051, device='cuda:0', grad_fn=<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(0.0029, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9195e-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(0.0005, 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(9.0708e-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(0.0004, 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(0.0015, 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(0.0006, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1102e-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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6217e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3433e-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(6.1887e-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(0.0007, 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(0.0004, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, 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(8.0172e-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(0.0011, 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(0.0007, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8133e-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(0.0023, 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(6.9395e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7408e-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(0.0002, 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(0.0015, 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(0.0007, 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(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1454e-05, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-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(5.6770e-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(0.0001, 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(0.0010, 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(0.0010, 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(0.0006, 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(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(0.0018, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4411e-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(0.0027, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5128e-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(0.0011, 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(0.0050, 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(0.0006, 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(0.0004, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7135e-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(0.0019, 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(0.0015, 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(8.1656e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5379e-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(7.1212e-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(0.0017, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7559e-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(0.0033, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-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(0.0001, 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(0.0013, 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(0.0008, 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(0.0051, 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(7.3634e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-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(0.0030, 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(0.0014, 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(0.0047, 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(0.0047, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8447e-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(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(9.8609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6276e-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(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2352e-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(0.0014, 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(0.0017, 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(7.2453e-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(0.0002, 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(0.0006, 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(0.0007, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0003, 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(0.0008, 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(0.0031, 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(0.0026, 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(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-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(0.0023, 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(0.0011, 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(0.0004, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-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(0.0023, 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(7.4511e-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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2373e-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(0.0004, 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(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1462e-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(0.0066, device='cuda:0', grad_fn=<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(0.0015, 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(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3311e-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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1494e-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(0.0025, 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(0.0005, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-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(8.5319e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5117e-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(0.0003, 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(0.0004, 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(0.0011, 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(7.2876e-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(0.0014, 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(7.0001e-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(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9800e-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(0.0023, 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(9.4947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-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(0.0019, 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(8.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5441e-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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0019, 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(0.0001, 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(7.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-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(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(0.0009, 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(0.0010, 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(0.0038, 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(0.0005, 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(0.0007, 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(0.0002, 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(0.0001, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-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(0.0027, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3452e-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(0.0014, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6540e-05, 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(7.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-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(5.2108e-05, 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(0.0003, 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(0.0007, 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(0.0001, 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(9.2040e-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(0.0010, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7216e-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(0.0003, 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(0.0021, 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(0.0010, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-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(0.0027, 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
				 {'precision': {'support': 0.8826754385964912, 'agreement': 0.5652173913043478, 'direct_attack': 0.24074074074074073, 'undercutter_attack': 0.4065040650406504, 'partial': 0.23333333333333334}, 'recall': {'support': 0.9284890426758939, 'agreement': 0.5148514851485149, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9050028105677347, 'agreement': 0.538860103626943, 'direct_attack': 0.2175732217573222, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09589041095890412}, 'support': {'support': 867, 'agreement': 202, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7040540540540541, 'recall': 0.7040540540540541, 'f1': 0.7040540540540541, 'support': None}, 'macro_avg': {'precision': 0.4656941938031127, 'recall': 0.46238294708286787, 'f1': 0.4490262849919369, 'support': None}, 'weighted_avg': {'precision': 0.6788674923493287, 'recall': 0.7040540540540541, 'f1': 0.6845355117677774, 'support': None}}
Loss: tensor(0.0017, 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(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(0.0011, 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(0.0019, 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(0.0037, 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(9.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1948e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-05, 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(7.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, 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(0.0008, 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(0.0013, 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(0.0007, 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(0.0007, 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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-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(0.0014, 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(0.0001, 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(8.4077e-05, device='cuda:0', grad_fn=<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(0.0026, 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(0.0015, 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(0.0014, 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(0.0013, 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(0.0008, 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(0.0007, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0575e-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(0.0027, 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(0.0001, 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(7.4844e-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(0.0014, 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(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-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(0.0039, 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(0.0028, device='cuda:0', grad_fn=<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(0.0007, 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(0.0001, 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(0.0008, 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(9.9245e-05, 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(0.0017, 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(0.0007, 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(0.0007, 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(0.0063, 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(4.2934e-05, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-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(0.0018, 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(0.0013, 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(0.0005, 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(0.0022, 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(0.0003, 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(0.0023, 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(0.0022, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6521e-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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-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(0.0037, device='cuda:0', grad_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(0.0005, 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(0.0018, 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(0.0001, 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(0.0004, 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(0.0073, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4276e-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(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-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(0.0027, 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(0.0024, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-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(7.5964e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0727e-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(9.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0977e-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(0.0023, 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(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(0.0002, 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(8.8559e-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(0.0013, 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(0.0041, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-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(0.0022, 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(8.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8044e-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(0.0001, 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(0.0005, 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(0.0017, 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(8.9646e-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(0.0012, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6742e-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(0.0004, 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(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(8.9345e-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(0.0008, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0009, 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(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8649e-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(0.0010, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1344e-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(6.6609e-05, device='cuda:0', grad_fn=<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(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(0.0004, 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(8.4017e-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(0.0014, 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(7.9082e-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(9.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1463e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8495e-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(6.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6751e-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(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(8.3714e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2101e-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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9434e-05, 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(0.0037, 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(0.0002, 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(0.0040, 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(9.6398e-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(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(5.7527e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0908e-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(7.5600e-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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4290e-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(0.0012, 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(0.0030, 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(0.0015, 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(7.6298e-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(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1251e-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(0.0003, 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(7.8386e-05, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1181e-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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7375e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7126e-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(0.0004, 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(0.0009, 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(0.0012, 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(6.1887e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, 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(0.0002, 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(8.7770e-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(0.0023, 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(0.0003, 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(0.0003, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6922e-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(6.8032e-05, device='cuda:0', grad_fn=<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(0.0059, 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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-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(0.0006, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, 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(0.0003, 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(5.9283e-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(8.5320e-05, device='cuda:0', grad_fn=<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(0.0009, 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(7.5873e-05, 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(0.0021, 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(0.0003, 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(0.0015, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, 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(7.4935e-05, 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(7.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8921e-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(0.0007, 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(0.0005, 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(0.0006, 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(0.0022, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7690e-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(0.0010, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-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(6.7700e-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(7.9688e-05, 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(0.0010, 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(0.0006, 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(0.0013, 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(0.0021, 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(0.0005, 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(0.0002, 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(0.0007, 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(0.0004, 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(3.5425e-05, device='cuda:0', grad_fn=<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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-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(0.0021, device='cuda:0', grad_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(0.0010, 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(0.0004, 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(0.0001, 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(0.0015, 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(0.0008, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2846e-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(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9819e-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(0.0005, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9698e-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(0.0009, 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(0.0002, 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(0.0029, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-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(0.0056, device='cuda:0', grad_fn=<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(8.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-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(5.7588e-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(7.9143e-05, 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(0.0030, 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(0.0002, 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(6.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1011e-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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5298e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-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(5.8647e-05, 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(0.0016, 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(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0706e-05, device='cuda:0', grad_fn=<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(8.2774e-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(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(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-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(0.0009, 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(6.8487e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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
				 {'precision': {'support': 0.8825466520307355, 'agreement': 0.5645161290322581, 'direct_attack': 0.24299065420560748, 'undercutter_attack': 0.4065040650406504, 'partial': 0.23333333333333334}, 'recall': {'support': 0.9284064665127021, 'agreement': 0.5172413793103449, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9048958919527293, 'agreement': 0.5398457583547558, 'direct_attack': 0.2184873949579832, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09589041095890412}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7040540540540541, 'recall': 0.7040540540540541, 'f1': 0.7040540540540541, 'support': None}, 'macro_avg': {'precision': 0.465978166728517, 'recall': 0.4628444106825955, 'f1': 0.44938486685463064, 'support': None}, 'weighted_avg': {'precision': 0.6786805972205631, 'recall': 0.7040540540540541, 'f1': 0.6844416674241944, 'support': None}}
Loss: tensor(5.0170e-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(0.0027, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, 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(0.0022, 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(0.0017, 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(0.0018, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5317e-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(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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7842e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8396e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0262e-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(0.0026, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(6.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-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(4.4448e-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(0.0023, 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(0.0003, 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(9.2615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1162e-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(0.0001, 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(0.0019, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-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(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0284e-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(0.0007, 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(0.0007, 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(6.9183e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-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(0.0014, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981e-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(9.3220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-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(0.0019, 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(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(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(0.0007, 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(0.0041, device='cuda:0', grad_fn=<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(0.0022, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-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(0.0015, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-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(0.0018, 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(0.0013, 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(8.3229e-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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5228e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, 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(8.8436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9223e-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(0.0006, 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(0.0005, 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(0.0009, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8477e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5885e-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(0.0014, 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(0.0010, 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(0.0010, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-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(0.0023, 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(7.8356e-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(0.0010, 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(0.0003, 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(9.1979e-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(0.0006, 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(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9104e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-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(6.6367e-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(6.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7630e-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(0.0027, 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(5.7285e-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(7.8992e-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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7750e-05, device='cuda:0', grad_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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-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(7.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7731e-05, 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(0.0005, 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(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-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(9.1312e-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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-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(7.1242e-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(7.9355e-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(8.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(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(0.0006, 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(0.0012, 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(4.9443e-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(0.0004, 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(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(0.0001, 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(0.0014, 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(0.0024, 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(9.2402e-05, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1433e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-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(0.0009, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6539e-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(9.2464e-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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-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(0.0015, 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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0666e-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(7.2029e-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(0.0003, 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(6.9304e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-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(0.0013, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-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(9.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3240e-05, 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(0.0008, 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(0.0001, 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(0.0001, 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(0.0019, 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(4.8687e-05, device='cuda:0', grad_fn=<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(6.0676e-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(0.0005, 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(0.0020, 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(0.0002, 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(6.0917e-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(0.0015, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5692e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-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(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0566e-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(7.8568e-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(8.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0004, 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(6.7760e-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(6.9576e-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(6.3673e-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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-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(0.0002, 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(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9395e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8525e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-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(9.1676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4226e-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(0.0012, 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(0.0008, 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(7.8628e-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(0.0022, 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(0.0039, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-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(0.0007, 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(0.0006, 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(5.7981e-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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7255e-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(4.8808e-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(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-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(6.6216e-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(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(0.0005, 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(0.0010, 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(0.0025, 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(7.4057e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9520e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3794e-05, device='cuda:0', grad_fn=<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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-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(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(5.8647e-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(0.0007, 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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4075e-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(0.0022, device='cuda:0', grad_fn=<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(7.0122e-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(0.0002, 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(0.0008, 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(0.0003, 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(0.0007, 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(0.0033, 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(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6318e-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(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-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(7.2543e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, 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(6.3189e-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(0.0007, 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(0.0005, 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(0.0006, 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(0.0006, 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(0.0009, 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(6.0918e-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(0.0006, 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(6.0403e-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(3.9331e-05, 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(0.0003, 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(5.2986e-05, device='cuda:0', grad_fn=<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(6.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, 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(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7136e-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(5.3106e-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(9.4975e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-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(0.0001, 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(6.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-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(0.0002, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-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(5.5650e-05, device='cuda:0', grad_fn=<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(3.7999e-05, 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(0.0002, 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>)
			------------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
				 {'precision': {'support': 0.8836443468715697, 'agreement': 0.5645161290322581, 'direct_attack': 0.24299065420560748, 'undercutter_attack': 0.4065040650406504, 'partial': 0.23333333333333334}, 'recall': {'support': 0.9284890426758939, 'agreement': 0.5198019801980198, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.905511811023622, 'agreement': 0.541237113402062, 'direct_attack': 0.2184873949579832, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09589041095890412}, 'support': {'support': 867, 'agreement': 202, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7047297297297297, 'recall': 0.7047297297297297, 'f1': 0.7047297297297297, 'support': None}, 'macro_avg': {'precision': 0.46619770569668384, 'recall': 0.46337304609276886, 'f1': 0.4497863216782704, 'support': None}, 'weighted_avg': {'precision': 0.6795385242138073, 'recall': 0.7047297297297297, 'f1': 0.6852390361320443, 'support': None}}
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8920e-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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5297e-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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8617e-05, device='cuda:0', grad_fn=<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(5.2895e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0004, 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(0.0016, 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(4.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9070e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-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(0.0002, 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(9.4613e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-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(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-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(0.0005, 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(0.0007, 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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9021e-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(0.0006, 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(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1775e-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(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(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4126e-05, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0015, 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(0.0004, 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(9.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6943e-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(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(8.8648e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-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(0.0005, 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(0.0003, 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(0.0001, 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(4.3418e-05, 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(0.0004, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8213e-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(0.0003, 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(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-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(3.8968e-05, 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(0.0004, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3644e-05, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0010e-05, 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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6165e-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(0.0008, 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(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-05, 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(0.0005, 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(9.8577e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8033e-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(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, 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(0.0016, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1645e-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(0.0015, 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(0.0009, 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(3.8453e-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(6.5580e-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(5.4863e-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(4.8384e-05, 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(0.0003, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-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(0.0001, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3330e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5259e-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(5.1956e-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(6.9546e-05, 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(0.0005, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-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(0.0016, 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(0.0006, 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(5.4802e-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(8.4986e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<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(3.9482e-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(0.0011, 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(8.7923e-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(0.0001, 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(0.0002, 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(7.7266e-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(0.0008, 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(5.7103e-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(0.0011, 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(6.6549e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-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(6.6277e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-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(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, 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(5.5771e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-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(0.0012, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-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(0.0021, 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(0.0003, 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(8.0869e-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(0.0001, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-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(0.0014, 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(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-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(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(6.0736e-05, device='cuda:0', grad_fn=<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(5.1744e-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(0.0012, 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(3.2065e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<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(7.7508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-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(8.4016e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9232e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-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(8.2927e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-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(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9163e-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(0.0023, 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(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5077e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969e-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(7.7993e-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(0.0001, 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(7.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9606e-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(7.0303e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1654e-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(7.9355e-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(0.0025, 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(0.0006, 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(0.0007, 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(0.0008, 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(6.6428e-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(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(0.0007, 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(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4290e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<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(5.0412e-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(0.0007, 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(0.0008, 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(0.0010, 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(9.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6105e-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(0.0003, 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(0.0010, 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(4.1692e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(8.5198e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9121e-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(0.0007, 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(6.6035e-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(5.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-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(3.7908e-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(8.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-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(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8848e-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(0.0002, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6375e-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(5.9828e-05, device='cuda:0', grad_fn=<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(8.1111e-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(5.6952e-05, 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(0.0005, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7751e-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(0.0004, 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(0.0001, 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(0.0004, 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(8.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6732e-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(0.0004, 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(0.0023, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8619e-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(5.0776e-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(0.0023, 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(5.1956e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-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(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0726e-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(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(4.2238e-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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-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(4.7142e-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(0.0017, 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(0.0005, 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(9.8820e-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(0.0004, 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(0.0001, 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(0.0006, 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
				 {'precision': {'support': 0.8846153846153846, 'agreement': 0.5608465608465608, 'direct_attack': 0.24761904761904763, 'undercutter_attack': 0.4048582995951417, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9284890426758939, 'agreement': 0.5221674876847291, 'direct_attack': 0.2, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9060213843556556, 'agreement': 0.5408163265306122, 'direct_attack': 0.22127659574468084, 'undercutter_attack': 0.48661800486618007, 'partial': 0.09655172413793105}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 130, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7054054054054054, 'recall': 0.7054054054054054, 'f1': 0.7054054054054054, 'support': None}, 'macro_avg': {'precision': 0.4678637206041924, 'recall': 0.46415149110156106, 'f1': 0.450256807127012, 'support': None}, 'weighted_avg': {'precision': 0.6806760997550471, 'recall': 0.7054054054054054, 'f1': 0.6858632194371148, 'support': None}}
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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7913e-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(7.5359e-05, 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(0.0003, 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(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6590e-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(6.0615e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8698e-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(6.4218e-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(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(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1454e-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(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-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(5.5953e-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(0.0017, 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(0.0007, 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(0.0016, 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(6.9182e-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(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6922e-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(5.1774e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7043e-05, device='cuda:0', grad_fn=<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(3.9392e-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(4.9565e-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(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-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(0.0005, 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(0.0015, 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(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(0.0028, 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(9.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2653e-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(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-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(9.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7549e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5895e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4025e-05, device='cuda:0', grad_fn=<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(4.2056e-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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0444e-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(5.6195e-05, device='cuda:0', grad_fn=<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(0.0018, 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(0.0017, 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(4.7536e-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(0.0007, 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(0.0019, 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(0.0007, 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(0.0003, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5953e-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(3.8695e-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(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-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(8.7832e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2262e-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(5.7557e-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(0.0015, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-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(0.0001, 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(0.0012, 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(0.0014, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-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(0.0007, 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(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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, 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(0.0010, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-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(9.8727e-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(5.0503e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, 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(5.9556e-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(9.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, 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(0.0001, 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(0.0003, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6832e-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(5.8435e-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(0.0007, 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(5.3712e-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(7.9263e-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(0.0023, 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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9760e-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(5.4954e-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(9.2061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-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(3.6969e-05, device='cuda:0', grad_fn=<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(0.0013, 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(9.2038e-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(6.4853e-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(8.6166e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0005, 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(0.0010, 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(6.4551e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3989e-05, device='cuda:0', grad_fn=<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(5.6770e-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(0.0015, 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(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-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(4.4963e-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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7257e-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(0.0003, 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(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-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(6.1493e-05, 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(0.0022, 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(4.2964e-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(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-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(0.0007, 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(5.1260e-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(9.0280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3775e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-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(0.0003, 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(7.5085e-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(7.8901e-05, device='cuda:0', grad_fn=<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(0.0016, 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(0.0010, 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(0.0009, 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(5.2501e-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(0.0009, 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(0.0003, 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(7.7448e-05, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-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(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-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(0.0010, 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(0.0004, 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(4.6446e-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(0.0003, 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(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-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(0.0008, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0038, 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(0.0001, 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(0.0004, 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(0.0029, 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(0.0001, 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(4.8505e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4117e-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(4.9080e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0004, 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(4.2904e-05, device='cuda:0', grad_fn=<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(8.4230e-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(4.0603e-05, 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(6.6427e-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(4.2328e-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(5.6074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-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(3.9694e-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(0.0043, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-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(0.0006, 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(0.0003, 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(0.0016, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-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(3.8241e-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(6.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8960e-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(4.6264e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2109e-05, device='cuda:0', grad_fn=<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(0.0006, 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(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(4.5931e-05, 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(0.0003, 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(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4733e-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(0.0005, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-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(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(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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4780e-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(5.0140e-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(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2976e-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(6.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-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(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(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-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(0.0005, 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(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-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(0.0003, 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(4.5629e-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(0.0006, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1533e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-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(7.0787e-05, 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(4.4781e-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(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9295e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1684e-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(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4652e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8855885588558856, 'agreement': 0.5631578947368421, 'direct_attack': 0.25, 'undercutter_attack': 0.4048582995951417, 'partial': 0.23333333333333334}, 'recall': {'support': 0.9295612009237876, 'agreement': 0.5270935960591133, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9070422535211267, 'agreement': 0.5445292620865141, 'direct_attack': 0.22127659574468084, 'undercutter_attack': 0.48661800486618007, 'partial': 0.09589041095890412}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.706081081081081, 'recall': 0.706081081081081, 'f1': 0.706081081081081, 'support': None}, 'macro_avg': {'precision': 0.4673876173042405, 'recall': 0.4650458009145663, 'f1': 0.45107130543548113, 'support': None}, 'weighted_avg': {'precision': 0.6807122786493554, 'recall': 0.706081081081081, 'f1': 0.6864553420707417, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(5.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-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(3.6606e-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(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0011, 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(0.0003, 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(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3116e-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(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-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(4.5659e-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(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(0.0004, 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(6.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0665e-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(6.5366e-05, 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(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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8850e-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(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3876e-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(4.8596e-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(4.8808e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-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(0.0007, 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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0311e-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(0.0007, 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(6.6821e-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(3.2791e-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(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, 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(7.8445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<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(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-05, device='cuda:0', grad_fn=<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(0.0009, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-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(0.0013, 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(0.0007, 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(0.0006, 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(0.0004, 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(8.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-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(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-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(7.2270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8476e-05, device='cuda:0', grad_fn=<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(0.0004, 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(5.2562e-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(0.0010, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-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(0.0003, 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(0.0003, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-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(4.5629e-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(0.0004, 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(0.0019, 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(0.0006, 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(7.5601e-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(0.0010, 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(4.5386e-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(8.8196e-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(6.6185e-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(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(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-05, device='cuda:0', grad_fn=<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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-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(4.1329e-05, device='cuda:0', grad_fn=<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(3.5486e-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(4.2298e-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(3.5395e-05, device='cuda:0', grad_fn=<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(3.9906e-05, device='cuda:0', grad_fn=<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(0.0005, 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(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8183e-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(5.5256e-05, 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(0.0018, 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(0.0018, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4771e-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(4.6597e-05, 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(6.5671e-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(0.0008, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4217e-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(0.0018, 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(7.3512e-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(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0002, 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(6.6549e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<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(0.0002, 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(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1746e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0007, 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(0.0026, 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(4.5356e-05, 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(0.0006, 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(5.7679e-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(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-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(7.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9061e-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(0.0003, 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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4644e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8579e-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(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-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(8.2805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(4.0330e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0647e-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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, 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(0.0010, 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(5.9434e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(5.5317e-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(6.1492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-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(0.0003, 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(0.0012, 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(0.0012, 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(0.0048, device='cuda:0', grad_fn=<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(5.6861e-05, device='cuda:0', grad_fn=<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(0.0016, 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(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9698e-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(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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7830e-05, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-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(8.4923e-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(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-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(7.8082e-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(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0007, 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(0.0010, 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(0.0003, 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(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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6546e-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(9.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2967e-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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-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(5.9041e-05, 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(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(0.0006, 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(0.0001, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-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(0.0009, 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(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-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(4.3539e-05, device='cuda:0', grad_fn=<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(5.1896e-05, device='cuda:0', grad_fn=<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(5.6800e-05, device='cuda:0', grad_fn=<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(4.3842e-05, device='cuda:0', grad_fn=<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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-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(4.5144e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-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(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<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(4.0875e-05, 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(0.0007, 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(3.2458e-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(4.6809e-05, device='cuda:0', grad_fn=<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(0.0008, 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(4.7506e-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(0.0026, 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(0.0017, 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(0.0002, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-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(3.9755e-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(0.0006, 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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9764e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-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(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0462e-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(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-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(6.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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>)
			------------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
				 {'precision': {'support': 0.8857142857142857, 'agreement': 0.5654450261780105, 'direct_attack': 0.24761904761904763, 'undercutter_attack': 0.40816326530612246, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9296424452133795, 'agreement': 0.5320197044334976, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.06086956521739131}, 'f1': {'support': 0.9071468767585819, 'agreement': 0.5482233502538071, 'direct_attack': 0.22033898305084748, 'undercutter_attack': 0.48899755501222486, 'partial': 0.09722222222222222}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 115}, 'micro_avg': {'precision': 0.7074324324324325, 'recall': 0.7074324324324325, 'f1': 0.7074324324324325, 'support': None}, 'macro_avg': {'precision': 0.4696641870324587, 'recall': 0.4661522189735985, 'f1': 0.4523857974595368, 'support': None}, 'weighted_avg': {'precision': 0.6823210253151191, 'recall': 0.7074324324324325, 'f1': 0.6878555700056992, 'support': None}}
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(0.0020, 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(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<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(6.0675e-05, 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(0.0003, 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(5.7254e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6055e-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(0.0001, 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(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(5.0140e-05, 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(2.9521e-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(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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0607e-05, device='cuda:0', grad_fn=<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(7.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7436e-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(0.0007, 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(4.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4784e-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(5.3742e-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(0.0006, 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(6.9697e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-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(0.0006, 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(2.6009e-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(6.9151e-05, 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(0.0006, 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(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7859e-05, 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(3.3488e-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(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(7.0000e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-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(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(0.0009, 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(9.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-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(0.0006, 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(0.0004, 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(5.3228e-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(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(0.0004, 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(6.5762e-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(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-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(0.0003, 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(7.8476e-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(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(3.8907e-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(5.7255e-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(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5467e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-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(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(2.4859e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5580e-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(9.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-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(0.0008, 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(5.8556e-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(0.0010, 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(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(0.0004, 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(0.0002, 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(0.0006, 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(0.0006, 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(0.0005, 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(0.0002, 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(3.2822e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6049e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-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(0.0002, 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(8.5016e-05, 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(0.0002, 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(5.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2736e-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(0.0013, 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(0.0006, 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(0.0002, 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(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(0.0004, 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(9.4278e-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(7.5298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(2.8250e-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(0.0009, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0005, 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(0.0003, 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(0.0002, 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(4.5992e-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(0.0012, 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(0.0006, 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(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5739e-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(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3148e-05, device='cuda:0', grad_fn=<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(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-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(0.0029, 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(3.4880e-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(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1412e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1573e-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(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5329e-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(4.2873e-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(8.2487e-05, 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(0.0004, 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(0.0011, 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(2.9491e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0100e-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(0.0015, 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(3.8604e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2939e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-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(0.0001, 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(0.0012, 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(0.0008, 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(0.0004, 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(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.1126e-05, 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(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(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(0.0001, 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(6.2067e-05, 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(0.0001, 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(0.0006, 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(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-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(0.0003, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7519e-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(7.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-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(9.1760e-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(3.6001e-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(1.8288e-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(0.0002, 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(0.0002, 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(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2219e-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(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(9.4401e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2824e-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(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8909e-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(9.2339e-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(9.3976e-05, device='cuda:0', grad_fn=<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(0.0006, 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(2.8280e-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(4.3570e-05, 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(0.0002, 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(2.7493e-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(0.0014, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, 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(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969e-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(0.0003, 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(3.7181e-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(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0005, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, 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(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-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(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-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(0.0003, 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(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-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(6.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0614e-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(6.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, 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(0.0005, 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(0.0002, 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(0.0004, 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(0.0015, 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(7.1090e-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(4.4751e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-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(7.0923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-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(0.0004, 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(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2441e-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(3.1005e-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(8.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-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(4.5447e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3499e-05, device='cuda:0', grad_fn=<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(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(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-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(0.0009, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-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(0.0024, 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(0.0004, 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(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<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(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4084e-05, device='cuda:0', grad_fn=<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(3.1187e-05, device='cuda:0', grad_fn=<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(2.4314e-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(3.1187e-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(0.0009, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8254e-05, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8865638766519823, 'agreement': 0.5677083333333334, 'direct_attack': 0.24761904761904763, 'undercutter_attack': 0.4065040650406504, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9295612009237876, 'agreement': 0.5369458128078818, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9075535512965051, 'agreement': 0.5518987341772152, 'direct_attack': 0.22033898305084748, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09655172413793105}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7074324324324325, 'recall': 0.7074324324324325, 'f1': 0.7074324324324325, 'support': None}, 'macro_avg': {'precision': 0.4699549265979682, 'recall': 0.46701624426432, 'f1': 0.452829574142256, 'support': None}, 'weighted_avg': {'precision': 0.6825093721297603, 'recall': 0.7074324324324325, 'f1': 0.6878656927300062, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<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(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(3.4911e-05, device='cuda:0', grad_fn=<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(0.0012, 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(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<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(5.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, 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(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, 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(4.0451e-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(2.2406e-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(0.0011, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0004, 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(6.9061e-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(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(3.9119e-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(4.4418e-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(0.0003, 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(0.0002, 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(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9567e-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(2.4677e-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(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(7.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6285e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-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(0.0006, 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(0.0006, 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(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(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(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-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(0.0011, 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(5.0503e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-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(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7901e-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(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0877e-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(0.0010, 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(6.3430e-05, 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(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(6.1492e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-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(0.0002, 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(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, 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(2.4707e-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(3.2882e-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(0.0006, 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(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(0.0001, 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(0.0004, 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(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-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(2.2376e-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(0.0008, 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(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0627e-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(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<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(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-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(3.7908e-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(0.0009, 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(2.9673e-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(0.0003, 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(7.9385e-05, device='cuda:0', grad_fn=<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(7.8598e-05, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-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(0.0001, 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(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(6.5852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7598e-05, device='cuda:0', grad_fn=<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(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9393e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-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(4.3327e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-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(5.9222e-05, device='cuda:0', grad_fn=<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(0.0009, 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(0.0001, 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(3.8453e-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(0.0003, 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(0.0002, 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(7.2996e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-05, 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(4.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-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(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<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(2.8643e-05, device='cuda:0', grad_fn=<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(6.3732e-05, 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(0.0017, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-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(3.4305e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1825e-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(9.8935e-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(0.0005, 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(0.0005, 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(0.0004, 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(0.0025, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-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(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(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<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(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, 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(7.9670e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2078e-05, 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(0.0004, 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(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-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(0.0010, 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(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(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0858e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-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(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-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(7.1998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9312e-05, device='cuda:0', grad_fn=<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(3.4759e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-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(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9556e-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(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5247e-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(0.0003, 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(3.8090e-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(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7284e-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(5.1320e-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(7.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-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(3.7575e-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(0.0005, 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(9.9757e-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(3.6636e-05, 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(4.2086e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(4.4902e-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(0.0017, 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(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(0.0004, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, 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(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(5.1502e-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(0.0002, 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(2.4526e-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(2.3254e-05, device='cuda:0', grad_fn=<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(0.0005, 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(6.6547e-05, device='cuda:0', grad_fn=<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(9.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4457e-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(9.6369e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, 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(0.0006, 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(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(9.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-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(3.1641e-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(5.4560e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3441e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3157e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, 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(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, 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(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-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(6.8910e-05, 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(0.0002, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4742e-05, device='cuda:0', grad_fn=<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(2.5888e-05, 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(3.9573e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4490e-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(0.0013, 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(0.0004, 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(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(4.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2369e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-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(0.0003, 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>)
			------------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
				 {'precision': {'support': 0.8865638766519823, 'agreement': 0.5677083333333334, 'direct_attack': 0.24528301886792453, 'undercutter_attack': 0.4065040650406504, 'partial': 0.25}, 'recall': {'support': 0.9295612009237876, 'agreement': 0.5369458128078818, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9075535512965051, 'agreement': 0.5518987341772152, 'direct_attack': 0.21940928270042193, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09722222222222222}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7074324324324325, 'recall': 0.7074324324324325, 'f1': 0.7074324324324325, 'support': None}, 'macro_avg': {'precision': 0.47121185877877814, 'recall': 0.46701624426432, 'f1': 0.452777733689029, 'support': None}, 'weighted_avg': {'precision': 0.6829782776930055, 'recall': 0.7074324324324325, 'f1': 0.6878359542380279, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3002e-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(0.0004, 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(8.4183e-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(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(5.6951e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0004, 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(0.0010, 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(0.0003, 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(0.0001, 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(7.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0015, 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(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(2.6887e-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(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-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(2.7644e-05, 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(2.6827e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6650e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7508e-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(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1642e-05, device='cuda:0', grad_fn=<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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-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(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0492e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4206e-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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, 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(0.0005, 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(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0004, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-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(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, 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(0.0001, 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(3.3881e-05, 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(3.7121e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6537e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<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(2.6403e-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(8.3540e-05, device='cuda:0', grad_fn=<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(3.0672e-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(0.0017, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<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(7.0301e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, 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(0.0012, 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(0.0007, 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(0.0006, 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(0.0011, 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(0.0002, 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(8.4272e-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(3.6606e-05, device='cuda:0', grad_fn=<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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4308e-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(0.0003, 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(2.1316e-05, 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(9.6224e-05, 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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-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(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0524e-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(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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1886e-05, device='cuda:0', grad_fn=<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(4.4539e-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(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6076e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-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(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<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(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9243e-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(0.0007, 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(5.5831e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, device='cuda:0', grad_fn=<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(3.0157e-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(2.5676e-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(2.7099e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7052e-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(3.6303e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-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(7.8356e-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(3.2488e-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(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-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(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(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-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(6.8940e-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(0.0002, 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(3.1429e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5700e-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(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(3.7121e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-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(0.0008, 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(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3744e-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(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-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(0.0006, 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(0.0005, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.4517e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0415e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8623e-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(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(0.0001, 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(7.6714e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3926e-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(0.0010, 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(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-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(6.2525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-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(5.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5235e-05, 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(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-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(3.0551e-05, device='cuda:0', grad_fn=<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(8.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-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(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-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(2.6191e-05, device='cuda:0', grad_fn=<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(3.0217e-05, device='cuda:0', grad_fn=<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(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-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(4.3509e-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(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4116e-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(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5241e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0383e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, 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(0.0003, 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(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3598e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-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(7.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1142e-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(4.4417e-05, device='cuda:0', grad_fn=<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(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(3.3397e-05, device='cuda:0', grad_fn=<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(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(6.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-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(7.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5541e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<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(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(6.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9221e-05, device='cuda:0', grad_fn=<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(4.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2644e-05, 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(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0427e-05, device='cuda:0', grad_fn=<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(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, 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(2.8795e-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(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-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(8.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0764e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, 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(0.0004, 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(0.0005, 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(0.0004, 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(6.0856e-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(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(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-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(4.8626e-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(2.2134e-05, device='cuda:0', grad_fn=<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(2.7735e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-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(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-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(4.0118e-05, device='cuda:0', grad_fn=<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(3.4608e-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(0.0002, 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(0.0004, 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(8.8892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8064e-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(0.0004, 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(0.0009, 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(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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
				 {'precision': {'support': 0.8865638766519823, 'agreement': 0.5654450261780105, 'direct_attack': 0.24761904761904763, 'undercutter_attack': 0.4048582995951417, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9295612009237876, 'agreement': 0.5320197044334976, 'direct_attack': 0.1984732824427481, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9075535512965051, 'agreement': 0.5482233502538071, 'direct_attack': 0.22033898305084748, 'undercutter_attack': 0.48661800486618007, 'partial': 0.09655172413793105}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7067567567567568, 'recall': 0.7067567567567568, 'f1': 0.7067567567567568, 'support': None}, 'macro_avg': {'precision': 0.4691731120778019, 'recall': 0.4660310225894432, 'f1': 0.45185712272105416, 'support': None}, 'weighted_avg': {'precision': 0.682016563423278, 'recall': 0.7067567567567568, 'f1': 0.6872300507446019, 'support': None}}
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(0.0005, 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(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-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(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2480e-05, device='cuda:0', grad_fn=<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(2.6917e-05, device='cuda:0', grad_fn=<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(6.9001e-05, device='cuda:0', grad_fn=<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(5.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-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(3.1610e-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(0.0010, 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(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-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(7.9203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-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(3.5092e-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(9.8102e-05, 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(0.0010, 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(0.0006, 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(3.6152e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7364e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, 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(0.0002, 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(0.0006, 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(0.0002, 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(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-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(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(0.0009, 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(2.5706e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7324e-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(0.0007, 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(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9128e-05, device='cuda:0', grad_fn=<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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-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(0.0001, 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(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-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(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0714e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<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(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7718e-05, 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(1.9621e-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(0.0008, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(9.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6619e-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(9.8684e-05, 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(2.0287e-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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-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(0.0005, 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(2.9007e-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(1.9287e-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(3.4911e-05, 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(0.0005, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5518e-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(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(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(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, 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(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(6.2860e-05, 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(7.9536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-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(0.0009, 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(3.1065e-05, device='cuda:0', grad_fn=<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(3.4093e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7840e-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(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(0.0002, 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(9.7514e-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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1664e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5239e-05, 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(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(0.0007, 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(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8122e-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(4.9019e-05, device='cuda:0', grad_fn=<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(0.0002, 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(9.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2481e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8384e-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(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-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(3.4365e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-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(2.8098e-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(2.7402e-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(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-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(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, 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(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(6.7153e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, 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(0.0003, 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(3.4275e-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(6.8366e-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(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7661e-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(0.0006, 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(0.0002, 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(1.8682e-05, 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(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, 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(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3013e-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(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6183e-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(7.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-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(0.0004, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-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(9.4259e-05, device='cuda:0', grad_fn=<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(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-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(3.9028e-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(0.0002, 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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-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(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-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(0.0007, 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(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(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6962e-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(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8286e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<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(2.8340e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-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(0.0009, 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(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(3.4614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-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(0.0002, 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(4.9110e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(9.2459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2689e-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(0.0011, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-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(0.0002, 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(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, 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(9.5207e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5034e-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(0.0003, 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(4.6173e-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(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0684e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-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(0.0008, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-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(3.0066e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-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(3.1156e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1279e-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(0.0004, 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(9.2386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4824e-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(9.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7488e-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(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7202e-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(8.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_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(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, 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(0.0001, 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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-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(3.8271e-05, 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(8.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-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(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, 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(4.0815e-05, device='cuda:0', grad_fn=<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(1.9136e-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(3.0520e-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(9.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, 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(2.5131e-05, device='cuda:0', grad_fn=<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(2.7220e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7353e-05, device='cuda:0', grad_fn=<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(8.4350e-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(1.5715e-05, 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(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(0.0002, device='cuda:0', grad_fn=<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(2.9763e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8868131868131868, 'agreement': 0.5684210526315789, 'direct_attack': 0.25961538461538464, 'undercutter_attack': 0.4008097165991903, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9307958477508651, 'agreement': 0.5320197044334976, 'direct_attack': 0.20610687022900764, 'undercutter_attack': 0.6073619631901841, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9082723691615082, 'agreement': 0.5496183206106869, 'direct_attack': 0.2297872340425532, 'undercutter_attack': 0.48292682926829267, 'partial': 0.09655172413793105}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 163, 'partial': 116}, 'micro_avg': {'precision': 0.7081081081081081, 'recall': 0.7081081081081081, 'f1': 0.7081081081081082, 'support': None}, 'macro_avg': {'precision': 0.4714077302008336, 'recall': 0.46732584263795224, 'f1': 0.45343129544419447, 'support': None}, 'weighted_avg': {'precision': 0.6835122336767073, 'recall': 0.7081081081081081, 'f1': 0.6885566648495293, 'support': None}}
Loss: tensor(3.8544e-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(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7057e-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(5.0277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9220e-05, 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(0.0003, 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(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-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(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(5.9918e-05, 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(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4159e-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(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-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(8.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, 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(0.0002, 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(4.0936e-05, device='cuda:0', grad_fn=<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(5.3590e-05, device='cuda:0', grad_fn=<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(4.8505e-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(2.5616e-05, device='cuda:0', grad_fn=<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(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-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(0.0003, 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(6.7699e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7413e-05, device='cuda:0', grad_fn=<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(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-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(2.2406e-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(0.0005, 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(5.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9978e-05, 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(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2859e-05, device='cuda:0', grad_fn=<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(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7251e-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(0.0004, 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(0.0004, 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(0.0006, 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(0.0003, 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(4.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, 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(0.0005, 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(0.0004, 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(5.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-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(0.0006, 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(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-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(0.0003, 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(0.0006, 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(4.4477e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<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(2.7250e-05, device='cuda:0', grad_fn=<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(9.9211e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, 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(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(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, 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(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-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(0.0002, 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(0.0004, 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(7.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-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(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, 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(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(8.1835e-05, 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(0.0006, 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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-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(5.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8373e-05, 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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-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(7.6964e-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(0.0006, 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(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(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(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-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(2.2497e-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(5.3469e-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(0.0006, 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(5.5043e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6883e-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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, 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(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(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3429e-05, 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(0.0002, 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(0.0005, 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(5.7102e-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(2.1498e-05, device='cuda:0', grad_fn=<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(2.2557e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, 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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8191e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3401e-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(1.9651e-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(3.0944e-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(0.0013, 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(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0001, 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(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(8.9775e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-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(0.0002, 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(8.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-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(3.1005e-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(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-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(2.4798e-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(1.9257e-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(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1349e-05, device='cuda:0', grad_fn=<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(3.1731e-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(1.9832e-05, 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(0.0002, 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(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(4.9564e-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(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(1.8137e-05, device='cuda:0', grad_fn=<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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-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(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-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(9.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-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(0.0006, 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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-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(9.7298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, 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(0.0003, 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(6.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4919e-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(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-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(2.2921e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<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(1.7864e-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(0.0001, 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(0.0008, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9393e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, 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(2.7917e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6121e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(1.9136e-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(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-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(0.0005, 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(9.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2551e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7583e-05, device='cuda:0', grad_fn=<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(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-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(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(7.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7547e-05, device='cuda:0', grad_fn=<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(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6159e-05, 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(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(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5800e-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(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-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(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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5044e-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(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9271e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<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(3.7757e-05, 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(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-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(2.6130e-05, device='cuda:0', grad_fn=<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(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1822e-05, 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(0.0009, 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(0.0002, 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(0.0004, 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(3.9180e-05, 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(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(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9365e-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(9.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<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(7.3298e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-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(7.6685e-05, 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(0.0002, 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
				 {'precision': {'support': 0.8868131868131868, 'agreement': 0.5654450261780105, 'direct_attack': 0.25961538461538464, 'undercutter_attack': 0.4065040650406504, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9307958477508651, 'agreement': 0.5320197044334976, 'direct_attack': 0.20610687022900764, 'undercutter_attack': 0.6097560975609756, 'partial': 0.06086956521739131}, 'f1': {'support': 0.9082723691615082, 'agreement': 0.5482233502538071, 'direct_attack': 0.2297872340425532, 'undercutter_attack': 0.48780487804878053, 'partial': 0.09722222222222222}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 115}, 'micro_avg': {'precision': 0.7087837837837838, 'recall': 0.7087837837837838, 'f1': 0.7087837837837839, 'support': None}, 'macro_avg': {'precision': 0.47195139459841184, 'recall': 0.4679096170383475, 'f1': 0.45426201074577427, 'support': None}, 'weighted_avg': {'precision': 0.6838427540689906, 'recall': 0.7087837837837838, 'f1': 0.6892190320132976, 'support': None}}
Loss: tensor(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0002, 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(4.8471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<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(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-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(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-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(0.0004, 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(5.0471e-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(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1632e-05, device='cuda:0', grad_fn=<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(0.0003, 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(1.9075e-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(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-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(7.8777e-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(0.0008, 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(0.0004, 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(5.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3102e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5700e-05, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7700e-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(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-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(6.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-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(0.0002, 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(0.0001, 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(2.2800e-05, 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(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<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(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(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(3.3972e-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(7.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2761e-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(4.1753e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, 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(8.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0667e-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(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-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(7.8401e-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(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(2.7705e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-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(0.0002, 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(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(3.8150e-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(9.7255e-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(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5104e-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(0.0004, 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(0.0006, 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(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4607e-05, device='cuda:0', grad_fn=<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(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0816e-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(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(6.0010e-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(0.0001, 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(0.0005, 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(0.0005, 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(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(3.7241e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-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(1.7622e-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(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(1.9681e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.6260e-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(7.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2651e-05, 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(6.2521e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3084e-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(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6204e-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(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-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(2.8371e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-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(0.0004, 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(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, 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(1.7834e-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(0.0002, device='cuda:0', grad_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(1.6139e-05, device='cuda:0', grad_fn=<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(9.8331e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5909e-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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4020e-05, device='cuda:0', grad_fn=<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(4.2207e-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(2.4192e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5336e-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(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(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-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(4.6900e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6728e-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(8.1261e-05, device='cuda:0', grad_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(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(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-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(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, 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(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1489e-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(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(3.3457e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-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(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(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<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(0.0005, 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(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(7.6983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-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(0.0004, 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(0.0007, 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(1.5200e-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(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-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(4.0148e-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(1.6260e-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(9.3433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9292e-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(8.2281e-05, device='cuda:0', grad_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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3754e-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(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1917e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, 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(3.1913e-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(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7554e-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(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, device='cuda:0', grad_fn=<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(2.2921e-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(6.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9732e-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(0.0002, 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(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4036e-05, device='cuda:0', grad_fn=<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(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.8349e-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(2.0771e-05, 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(4.7081e-05, device='cuda:0', grad_fn=<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(0.0006, 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(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0853e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4432e-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(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(0.0002, 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(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3689e-05, device='cuda:0', grad_fn=<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(4.2783e-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(1.7350e-05, device='cuda:0', grad_fn=<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(7.4145e-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(6.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0407e-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(1.7319e-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(7.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6886e-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(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-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(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(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, 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(0.0003, 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(3.1852e-05, 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(0.0004, 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(1.6350e-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(7.3511e-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(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1997e-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(0.0005, 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(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0079e-05, 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(0.0003, 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(8.5982e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-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(4.1450e-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(0.0003, 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(4.1298e-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(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6910e-05, device='cuda:0', grad_fn=<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(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, 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(2.7886e-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(0.0002, 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
				 {'precision': {'support': 0.8876651982378855, 'agreement': 0.5654450261780105, 'direct_attack': 0.25961538461538464, 'undercutter_attack': 0.4032258064516129, 'partial': 0.2413793103448276}, 'recall': {'support': 0.930715935334873, 'agreement': 0.5320197044334976, 'direct_attack': 0.20610687022900764, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9086809470124013, 'agreement': 0.5482233502538071, 'direct_attack': 0.2297872340425532, 'undercutter_attack': 0.4854368932038835, 'partial': 0.09655172413793105}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7081081081081081, 'recall': 0.7081081081081081, 'f1': 0.7081081081081082, 'support': None}, 'macro_avg': {'precision': 0.47146614516554414, 'recall': 0.4677886870289122, 'f1': 0.4537360297301153, 'support': None}, 'weighted_avg': {'precision': 0.683541925426233, 'recall': 0.7081081081081081, 'f1': 0.6885951475400499, 'support': None}}
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-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(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8101e-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(0.0004, 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(6.3944e-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(1.6805e-05, 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(9.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<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(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(7.4357e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-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(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, 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(5.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_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(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9128e-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(3.5092e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0697e-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(2.9612e-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(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8265e-05, device='cuda:0', grad_fn=<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(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-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(0.0003, 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(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, 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(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4236e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4659e-05, device='cuda:0', grad_fn=<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(7.3450e-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(2.0862e-05, device='cuda:0', grad_fn=<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(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7445e-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(0.0008, 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(3.8907e-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(2.2860e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6720e-05, 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(7.0514e-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(1.2868e-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(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<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(2.5101e-05, device='cuda:0', grad_fn=<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(9.4838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1934e-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(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-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(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-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(4.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-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(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1944e-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(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(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3189e-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(2.9582e-05, 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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-05, device='cuda:0', grad_fn=<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(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-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(9.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, 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(7.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4326e-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(5.1683e-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(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8967e-05, 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(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, 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(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-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(0.0002, 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(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-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(5.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9735e-05, device='cuda:0', grad_fn=<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(1.3716e-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(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-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(2.7190e-05, 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(0.0003, 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(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-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(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<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(4.1360e-05, 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(0.0002, 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(0.0005, 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(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-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(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(1.9832e-05, 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(6.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, 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(9.8242e-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(6.6436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, 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(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-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(4.3479e-05, 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(0.0002, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6315e-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(1.9197e-05, device='cuda:0', grad_fn=<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(7.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9711e-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(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, 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(2.3436e-05, 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(6.0706e-05, device='cuda:0', grad_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(2.9885e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8649e-05, device='cuda:0', grad_fn=<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(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9605e-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(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6470e-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(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5678e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, 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(0.0002, 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(0.0001, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(0.0003, 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(2.6130e-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(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5544e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, 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(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4673e-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(0.0007, 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(8.3839e-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(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<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(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0801e-05, device='cuda:0', grad_fn=<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(8.6709e-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(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-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(2.1165e-05, 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(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4593e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-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(0.0002, 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(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(2.2255e-05, 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(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5109e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-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(1.2747e-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(1.7774e-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(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(7.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9949e-05, 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(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, 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(0.0003, 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(6.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-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(0.0005, 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(9.2220e-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(0.0003, 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(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1268e-05, device='cuda:0', grad_fn=<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(4.4114e-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(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6084e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-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(5.4268e-05, device='cuda:0', grad_fn=<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(2.0862e-05, device='cuda:0', grad_fn=<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(3.8635e-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(4.6839e-05, device='cuda:0', grad_fn=<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(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5911e-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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-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(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8798e-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(0.0012, 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(1.4927e-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(0.0004, 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(2.8522e-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(0.0004, 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(0.0003, 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(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(1.8228e-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(0.0002, 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(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2240e-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(4.3653e-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(0.0002, 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(2.6342e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-05, device='cuda:0', grad_fn=<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>)
			------------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
				 {'precision': {'support': 0.8877887788778878, 'agreement': 0.5677083333333334, 'direct_attack': 0.2621359223300971, 'undercutter_attack': 0.4048582995951417, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9307958477508651, 'agreement': 0.5369458128078818, 'direct_attack': 0.2076923076923077, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9087837837837839, 'agreement': 0.5518987341772152, 'direct_attack': 0.2317596566523605, 'undercutter_attack': 0.48661800486618007, 'partial': 0.09655172413793105}, 'support': {'support': 867, 'agreement': 203, 'direct_attack': 130, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7094594594594594, 'recall': 0.7094594594594594, 'f1': 0.7094594594594594, 'support': None}, 'macro_avg': {'precision': 0.47277412889625736, 'recall': 0.4691069786796474, 'f1': 0.4551223807234942, 'support': None}, 'weighted_avg': {'precision': 0.6847514148583184, 'recall': 0.7094594594594594, 'f1': 0.6899223592847133, 'support': None}}
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-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(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(6.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0007, 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(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8969e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-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(3.0066e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, 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(0.0002, 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(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(0.0002, 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(4.2056e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9814e-05, 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(0.0004, 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(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-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(0.0003, 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(0.0002, device='cuda:0', grad_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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-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(0.0003, 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(4.3221e-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(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-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(2.8310e-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(1.8954e-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(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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-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(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3662e-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(4.8417e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5270e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0002, 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(7.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2616e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0588e-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(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7129e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-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(1.7047e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-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(2.0771e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-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(3.8876e-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(8.7338e-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(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8396e-05, device='cuda:0', grad_fn=<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(1.8349e-05, device='cuda:0', grad_fn=<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(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2738e-05, device='cuda:0', grad_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(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<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(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1395e-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(2.7856e-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(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(2.0166e-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(2.0377e-05, device='cuda:0', grad_fn=<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(3.2125e-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(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-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>)
Loss: tensor(2.1922e-05, 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(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6164e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215e-05, device='cuda:0', grad_fn=<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(9.9725e-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(0.0002, 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(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(8.3896e-05, 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(0.0010, 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(8.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7955e-05, 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(0.0001, 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(2.2164e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5972e-05, 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(7.3783e-05, device='cuda:0', grad_fn=<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(1.9560e-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(8.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-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(2.7311e-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(0.0002, 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(0.0007, 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(8.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3055e-05, device='cuda:0', grad_fn=<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(9.5329e-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(3.5425e-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(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, 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(1.2535e-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(4.2937e-05, device='cuda:0', grad_fn=<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(8.1766e-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(2.8976e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-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(6.0856e-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(1.8591e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-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(3.6121e-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(0.0003, 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(1.0295e-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(1.2263e-05, 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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-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(7.0067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7857e-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(0.0002, 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(6.1732e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, 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(0.0003, 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(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-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(2.1710e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1068e-05, device='cuda:0', grad_fn=<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(0.0001, 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(3.9240e-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(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1108e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9651e-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(6.0414e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9523e-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(0.0015, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4929e-05, device='cuda:0', grad_fn=<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(1.6472e-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(5.0260e-05, device='cuda:0', grad_fn=<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(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1542e-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(7.9658e-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(3.9785e-05, 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(2.9400e-05, 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(3.7999e-05, device='cuda:0', grad_fn=<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(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1161e-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(2.4919e-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(8.1390e-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(0.0007, 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(1.6502e-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(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9993e-05, 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(0.0001, 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(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_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(2.6009e-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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-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(1.4080e-05, 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(0.0003, 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(1.8591e-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(7.9109e-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(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-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(0.0004, 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(0.0002, 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(2.9764e-05, 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(0.0005, 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(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, device='cuda:0', grad_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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(1.8621e-05, 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(0.0002, 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(1.1445e-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(0.0001, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2712e-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(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-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(9.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-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(6.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-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(0.0009, 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(7.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, 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(0.0002, 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(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0002, 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(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(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5680e-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(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, 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(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4026e-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(4.0512e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<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(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0024e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-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(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-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(0.0001, 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(2.2557e-05, 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(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(0.0002, 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(4.1329e-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(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(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8606e-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>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8877887788778878, 'agreement': 0.5677083333333334, 'direct_attack': 0.2647058823529412, 'undercutter_attack': 0.4032258064516129, 'partial': 0.2413793103448276}, 'recall': {'support': 0.9318706697459584, 'agreement': 0.5369458128078818, 'direct_attack': 0.20610687022900764, 'undercutter_attack': 0.6097560975609756, 'partial': 0.0603448275862069}, 'f1': {'support': 0.9092957746478872, 'agreement': 0.5518987341772152, 'direct_attack': 0.2317596566523605, 'undercutter_attack': 0.4854368932038835, 'partial': 0.09655172413793105}, 'support': {'support': 866, 'agreement': 203, 'direct_attack': 131, 'undercutter_attack': 164, 'partial': 116}, 'micro_avg': {'precision': 0.7094594594594594, 'recall': 0.7094594594594594, 'f1': 0.7094594594594594, 'support': None}, 'macro_avg': {'precision': 0.4729616222721205, 'recall': 0.46900485558600613, 'f1': 0.4549885565638555, 'support': None}, 'weighted_avg': {'precision': 0.6843752547440657, 'recall': 0.7094594594594594, 'f1': 0.689633614452663, '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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
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': 0.0871345029239766, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 1.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.16030123722431416, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.08703271028037383, 'recall': 0.08703271028037383, 'f1': 0.08703271028037383, 'support': None}, 'macro_avg': {'precision': 0.017426900584795322, 'recall': 0.2, 'f1': 0.032060247444862834, 'support': None}, 'weighted_avg': {'precision': 0.007583551948406843, 'recall': 0.08703271028037383, 'f1': 0.013951451136929213, 'support': None}}
Loss: tensor(1.8175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0252, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7389347336834209, 'agreement': 0.5932203389830508, 'direct_attack': 0.02127659574468085, 'undercutter_attack': 0.29304029304029305, 'partial': 0.0}, 'recall': {'support': 0.9544573643410853, 'agreement': 0.125, 'direct_attack': 0.006711409395973154, 'undercutter_attack': 0.5633802816901409, 'partial': 0.0}, 'f1': {'support': 0.8329809725158563, 'agreement': 0.20648967551622419, 'direct_attack': 0.01020408163265306, 'undercutter_attack': 0.3855421686746988, 'partial': 0.0}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.643107476635514, 'recall': 0.643107476635514, 'f1': 0.643107476635514, 'support': None}, 'macro_avg': {'precision': 0.3292943922902891, 'recall': 0.32990981108543993, 'f1': 0.28704337966788646, 'support': None}, 'weighted_avg': {'precision': 0.5686123098447567, 'recall': 0.643107476635514, 'f1': 0.5687621897756887, 'support': None}}
Loss: tensor(2.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3344, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9327024185068349, 'agreement': 0.6098654708520179, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.34226190476190477, 'partial': 0.07317073170731707}, 'recall': {'support': 0.8594961240310077, 'agreement': 0.4857142857142857, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.8098591549295775, 'partial': 0.027522935779816515}, 'f1': {'support': 0.8946041351487645, 'agreement': 0.540755467196819, 'direct_attack': 0.14838709677419354, 'undercutter_attack': 0.48117154811715485, 'partial': 0.04}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6799065420560748, 'recall': 0.6799065420560748, 'f1': 0.6799065420560748, 'support': None}, 'macro_avg': {'precision': 0.42017153373704347, 'recall': 0.467390983312414, 'f1': 0.42098364944738637, 'support': None}, 'weighted_avg': {'precision': 0.7074612980465074, 'recall': 0.6799065420560748, 'f1': 0.683083548797094, 'support': None}}
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0601, 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(3.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7936, 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(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, 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(1.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3163, 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(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8907882241215574, 'agreement': 0.5236486486486487, 'direct_attack': 0.4056603773584906, 'undercutter_attack': 0.3705179282868526, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9080348499515973, 'agreement': 0.5535714285714286, 'direct_attack': 0.2905405405405405, 'undercutter_attack': 0.6549295774647887, 'partial': 0.01834862385321101}, 'f1': {'support': 0.8993288590604027, 'agreement': 0.5381944444444444, 'direct_attack': 0.3385826771653543, 'undercutter_attack': 0.47328244274809167, 'partial': 0.034782608695652174}, 'support': {'support': 1033, 'agreement': 280, 'direct_attack': 148, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.7190420560747663, 'recall': 0.7190420560747663, 'f1': 0.7190420560747663, 'support': None}, 'macro_avg': {'precision': 0.5047897023497765, 'recall': 0.48508500407631316, 'f1': 0.456834206422789, 'support': None}, 'weighted_avg': {'precision': 0.7101579860620989, 'recall': 0.7190420560747663, 'f1': 0.7014070112689065, 'support': None}}
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, 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(0.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(1.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, 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(0.5646, 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(0.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0575, 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(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, 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(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, 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(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5517, 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(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2519, 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(1.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, 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(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, 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(0.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, 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(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, 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(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, 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(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(2.2206, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, 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(0.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, 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(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, 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(1.7361, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2153, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5452, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8190871369294606, 'agreement': 0.5869565217391305, 'direct_attack': 0.3617021276595745, 'undercutter_attack': 0.358974358974359, 'partial': 0.16417910447761194}, 'recall': {'support': 0.9563953488372093, 'agreement': 0.19285714285714287, 'direct_attack': 0.11409395973154363, 'undercutter_attack': 0.5915492957746479, 'partial': 0.2018348623853211}, 'f1': {'support': 0.8824318283415289, 'agreement': 0.2903225806451613, 'direct_attack': 0.17346938775510204, 'undercutter_attack': 0.4468085106382979, 'partial': 0.18106995884773663}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6799065420560748, 'recall': 0.6799065420560748, 'f1': 0.6799065420560748, 'support': None}, 'macro_avg': {'precision': 0.45817984995602723, 'recall': 0.41134612191717296, 'f1': 0.39482045324556536, 'support': None}, 'weighted_avg': {'precision': 0.6614540010408032, 'recall': 0.6799065420560748, 'f1': 0.6431018354145179, 'support': None}}
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6113, 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(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, 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(1.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1689, 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(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, 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(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2128, 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(1.1149, 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(0.0017, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3993, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1716, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, 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(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, 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(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, 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(0.0286, 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(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, 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(0.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, 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(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0291, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4667, 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(0.0060, 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(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, 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(0.0339, 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(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4086, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, 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(0.5543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5123, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8088, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, 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(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0009, 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(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, 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(3.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4498, 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(0.9005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, 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(0.0855, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, 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(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, 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(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9301, 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(1.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, 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(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, 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(0.0242, 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(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, 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(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, 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(0.1422, 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(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, 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(0.3937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, 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(0.0248, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9322974472807991, 'agreement': 0.480225988700565, 'direct_attack': 0.32608695652173914, 'undercutter_attack': 0.2660332541567696, 'partial': 0.11377245508982035}, 'recall': {'support': 0.813953488372093, 'agreement': 0.30357142857142855, 'direct_attack': 0.10067114093959731, 'undercutter_attack': 0.7887323943661971, 'partial': 0.1743119266055046}, 'f1': {'support': 0.8691153647180548, 'agreement': 0.37199124726477023, 'direct_attack': 0.15384615384615385, 'undercutter_attack': 0.3978685612788633, 'partial': 0.13768115942028986}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6255841121495327, 'recall': 0.6255841121495327, 'f1': 0.6255841121495327, 'support': None}, 'macro_avg': {'precision': 0.42368322034993866, 'recall': 0.43624807577096414, 'f1': 0.3861004973056264, 'support': None}, 'weighted_avg': {'precision': 0.6982237842562697, 'recall': 0.6255841121495327, 'f1': 0.6399020237293547, 'support': None}}
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, 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(1.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, 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(0.4312, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, 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(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7693, 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(0.0721, 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(0.0100, 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(0.0055, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, 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(0.2655, 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(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0025, 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(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, 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(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0011, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, 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(0.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, 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(0.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, 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(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, 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(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3748, 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(0.2779, device='cuda:0', grad_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(0.0058, 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(0.0089, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, 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(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5996, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0078, 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(0.0027, 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(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4190, 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(0.0088, 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(0.0135, 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(0.0218, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, 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(0.0005, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0010, 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(2.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0042, 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(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, 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(1.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2821, 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(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, 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(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7814, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2239, 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(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9322, 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(0.0112, 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(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, 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(1.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6121, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0899, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8543, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, 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(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0120, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5363, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, 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(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0045, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0006, 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(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, 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(0.0038, 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(0.0057, 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(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4136, 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(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6069, 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(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2943, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, 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(0.5187, 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(0.0444, 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(0.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, 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(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2591, 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(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, 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(0.2472, 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(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3433, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8642732049036778, 'agreement': 0.6363636363636364, 'direct_attack': 0.3543307086614173, 'undercutter_attack': 0.3605150214592275, 'partial': 0.21739130434782608}, 'recall': {'support': 0.9563953488372093, 'agreement': 0.425, 'direct_attack': 0.30201342281879195, 'undercutter_attack': 0.5915492957746479, 'partial': 0.045871559633027525}, 'f1': {'support': 0.9080036798528058, 'agreement': 0.5096359743040686, 'direct_attack': 0.32608695652173914, 'undercutter_attack': 0.44799999999999995, 'partial': 0.07575757575757575}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.7242990654205608, 'recall': 0.7242990654205608, 'f1': 0.7242990654205608, 'support': None}, 'macro_avg': {'precision': 0.486574775147157, 'recall': 0.4641659254127354, 'f1': 0.45349683728723783, 'support': None}, 'weighted_avg': {'precision': 0.6996470948914066, 'recall': 0.7242990654205608, 'f1': 0.7010621511054614, 'support': None}}
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9393, 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(0.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2558, 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(0.0011, 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(0.0007, 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(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7307, 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(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0204, 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(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, 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(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, 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(0.1941, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, 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(0.2841, 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(0.0024, 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(0.0228, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, 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(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, 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(0.0028, 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(0.0711, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0013, 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(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, 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(0.0200, 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(0.0049, 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(0.0297, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, 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(0.0016, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, 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(0.0004, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, 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(0.0024, 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(0.0158, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, 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(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0694, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, 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(0.0072, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, 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(0.0082, 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(3.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, 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(0.0109, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0016, 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(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, 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(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, 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(0.0240, 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(0.0006, 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(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1091, 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(0.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, 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(0.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, 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(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, 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(0.0192, 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(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, 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(0.0041, 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(0.0793, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, 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(0.0153, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, 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(0.0714, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, 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(0.2314, 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(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, 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(0.0212, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, 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(0.0017, 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(0.0185, 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(0.0105, device='cuda:0', grad_fn=<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(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<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(0.0203, 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(0.0626, 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(0.0016, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, 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(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2748, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, 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(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, 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(0.0035, 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(0.0026, 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(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, 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(0.0139, 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(0.0051, 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(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, 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(0.0156, 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(0.0021, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, 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(1.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, 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(0.0033, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0267, 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(0.0165, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2598, 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(0.0042, 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(0.0417, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8149, 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(0.0004, 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(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4289, 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(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2155, 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(0.0002, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, 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(0.0418, 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(0.0005, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.830997526793075, 'agreement': 0.6310679611650486, 'direct_attack': 0.38333333333333336, 'undercutter_attack': 0.3303030303030303, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9767441860465116, 'agreement': 0.23214285714285715, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.7676056338028169, 'partial': 0.01834862385321101}, 'f1': {'support': 0.8979955456570157, 'agreement': 0.33942558746736295, 'direct_attack': 0.2200956937799043, 'undercutter_attack': 0.46186440677966106, 'partial': 0.034782608695652174}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.7050233644859814, 'recall': 0.7050233644859814, 'f1': 0.7050233644859814, 'support': None}, 'macro_avg': {'precision': 0.5018070369855642, 'recall': 0.4298407433905559, 'f1': 0.39083276847591925, 'support': None}, 'weighted_avg': {'precision': 0.6861223756306642, 'recall': 0.7050233644859814, 'f1': 0.6565075210821528, 'support': None}}
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0908, 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(0.0016, 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(0.0152, 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(0.9131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3586, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, 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(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, 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(0.0128, 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(0.0077, 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(0.0128, 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(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(0.0204, 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(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5946e-05, 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(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3412e-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(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7413, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, 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(0.0002, 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(0.0011, 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(0.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9472, 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(0.0072, 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(7.4965e-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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, 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(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, 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(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9867, 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(0.0017, 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(0.0219, 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(0.0061, 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(0.0013, 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(0.0081, 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(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-05, 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(0.0827, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, 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(0.0023, 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(0.0175, 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(0.0201, 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(7.6449e-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(0.0007, 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(7.3058e-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(0.0260, 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(0.0101, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, 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(0.0005, 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(0.0102, 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(0.0006, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-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(0.0027, 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(0.0034, 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(0.0003, 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(9.0193e-05, 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(0.0139, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0003, 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(0.0007, 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(0.0036, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0004, 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(0.0001, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4713e-05, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0013, 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(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<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(0.0001, 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(0.0017, 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(0.0059, 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(0.0387, 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(0.0015, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, 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(0.0048, 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(0.0021, 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(0.0002, 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(0.0030, 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(0.0066, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<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(0.0042, 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(0.0012, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-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(0.0059, 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(0.0476, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3009e-05, 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(0.0006, 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(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8256e-05, 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(0.0025, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(7.9991e-05, 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(0.0037, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, 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(0.0009, 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(0.0002, 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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, 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(0.0018, 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(0.0018, 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(0.0245, 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(0.0031, 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(0.0002, 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(0.0008, 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(9.5976e-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(0.0028, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, 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(0.0002, 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(0.0010, 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(0.0001, 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(0.0018, 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(0.0044, 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(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0001, 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(0.0022, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, 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(0.0020, 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(0.0002, 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(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0081, 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(0.0060, 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(0.0002, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0015, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, 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(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, 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(0.0046, 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(0.0031, 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(0.0060, 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(0.0015, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7732e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, 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(0.0053, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, 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(0.0084, device='cuda:0', grad_fn=<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(0.0004, 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(0.0011, 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(0.0025, 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(0.0002, 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(0.0104, 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(0.0043, 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(0.0003, 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(0.0073, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, 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(0.0018, 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(0.0041, 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(0.0054, 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(0.0011, 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(0.0075, 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(0.0021, 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(0.0004, 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(0.0122, 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(0.0122, 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(0.0017, 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(0.0207, 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(0.0201, 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(0.0666, 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(0.0013, 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(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<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(0.0246, 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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8753437213565536, 'agreement': 0.5838509316770186, 'direct_attack': 0.26, 'undercutter_attack': 0.3205882352941177, 'partial': 0.15714285714285714}, 'recall': {'support': 0.9253875968992248, 'agreement': 0.3357142857142857, 'direct_attack': 0.087248322147651, 'undercutter_attack': 0.7676056338028169, 'partial': 0.10091743119266056}, 'f1': {'support': 0.8996702779086198, 'agreement': 0.42630385487528344, 'direct_attack': 0.1306532663316583, 'undercutter_attack': 0.45228215767634855, 'partial': 0.12290502793296089}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6904205607476636, 'recall': 0.6904205607476636, 'f1': 0.6904205607476636, 'support': None}, 'macro_avg': {'precision': 0.43938514909410936, 'recall': 0.4433746539513278, 'f1': 0.4063629169449742, 'support': None}, 'weighted_avg': {'precision': 0.6823744638725845, 'recall': 0.6904205607476636, 'f1': 0.6687575100963354, 'support': None}}
Loss: tensor(0.0085, 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(0.0020, 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(0.0020, 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(0.0077, 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(0.0057, 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(0.0002, 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(0.0026, 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(0.0004, 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(0.0012, 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(0.0027, 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(0.0045, 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(0.0013, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0023, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0009, 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(0.0046, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, 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(0.0013, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2331, 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(0.0032, 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(0.0016, 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(0.0039, 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(0.0024, 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(0.0019, 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(0.0001, 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(0.0002, 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(0.0010, 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(0.0042, 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(0.0223, 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(0.0003, 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(0.0047, 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(0.0020, device='cuda:0', grad_fn=<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(0.0001, 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(0.0117, 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(0.0030, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0041, 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(0.0089, 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(0.0044, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4916e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8478e-05, 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(0.0006, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(9.2797e-05, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0018, 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(0.0016, 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(0.0002, 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(0.0069, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<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(0.0002, 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(0.0010, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2724, 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(0.0128, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, 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(0.0004, 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(0.0002, 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(0.0106, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3251e-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(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0003, 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(0.0038, 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(0.0002, 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(8.2049e-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(0.0029, 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(0.0013, 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(0.0002, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8458e-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(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-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(0.0013, 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(0.0015, 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(0.0018, 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(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(6.6791e-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(0.0069, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9940e-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(0.0069, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, 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(0.0008, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, 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(8.9618e-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(0.0017, 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(0.0007, 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(0.0012, 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(0.0002, 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(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0020, 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(0.0065, 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(0.0014, 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(0.0021, 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(8.7741e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.0001, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7468e-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(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(0.0005, 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(0.0074, 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(0.0035, 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(0.0024, 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(0.0024, 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(0.0002, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, 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(0.0004, 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(0.0004, 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(0.0043, device='cuda:0', grad_fn=<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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7368e-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(0.0022, 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(0.0029, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6187e-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(0.0015, 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(0.0022, 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(0.0008, 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(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3644e-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(0.0031, 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(0.0028, 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(0.0008, 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(0.0023, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8790e-05, 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(0.0002, 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(0.0013, 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(0.0011, 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(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9022e-05, 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(0.0011, 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(0.0089, device='cuda:0', grad_fn=<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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3352e-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(0.0076, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3553e-05, device='cuda:0', grad_fn=<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(8.5108e-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(0.0018, 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(0.0034, 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(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(0.0001, 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(0.0162, 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(0.0017, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8699e-05, 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(0.0066, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6923e-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(0.0044, 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(0.0022, 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(0.0024, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<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(0.0007, 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(0.0064, 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(8.8981e-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(0.0009, 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(7.5692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5510e-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(0.0009, 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(0.0035, 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(0.0004, 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(0.0033, 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(0.0040, 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(0.0036, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.0001, 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(0.0019, 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(0.0014, 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(0.0023, 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(0.0032, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0008, 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(0.0023, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0727e-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(0.0064, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0021, 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(0.0025, 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(0.0026, 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(0.0010, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0002, 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(0.0084, 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(0.0004, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1192e-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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, 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(0.0015, 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(0.0180, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9244e-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(0.0012, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8435609334485739, 'agreement': 0.6, 'direct_attack': 0.3698630136986301, 'undercutter_attack': 0.34265734265734266, 'partial': 0.1568627450980392}, 'recall': {'support': 0.9457364341085271, 'agreement': 0.3107142857142857, 'direct_attack': 0.18120805369127516, 'undercutter_attack': 0.6901408450704225, 'partial': 0.07339449541284404}, 'f1': {'support': 0.8917313841936957, 'agreement': 0.4094117647058823, 'direct_attack': 0.24324324324324323, 'undercutter_attack': 0.45794392523364486, 'partial': 0.1}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6985981308411215, 'recall': 0.6985981308411215, 'f1': 0.6985981308411215, 'support': None}, 'macro_avg': {'precision': 0.46258880698051713, 'recall': 0.44023882279947096, 'f1': 0.42046606347529325, 'support': None}, 'weighted_avg': {'precision': 0.6772312232669702, 'recall': 0.6985981308411215, 'f1': 0.6700194878691367, 'support': None}}
Loss: tensor(7.6540e-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(0.0034, 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(0.0037, 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(0.0062, 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(0.0015, 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(0.0022, 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(0.0004, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-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(0.0022, 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(0.0017, 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(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8820e-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(8.2776e-05, 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(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9839e-05, 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(0.0016, 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(0.0001, 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(0.0013, 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(0.0439, 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(0.0008, 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(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(0.0011, 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(0.0042, 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(0.0028, 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(0.0072, 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(7.9930e-05, 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(8.1898e-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(0.0001, 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(0.0116, device='cuda:0', grad_fn=<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(0.0022, 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(0.0033, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(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(0.0024, 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(0.0018, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-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(0.0001, 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(0.0002, 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(0.0008, 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(0.0033, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5268e-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(0.0024, 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(0.0003, 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(0.0007, 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(0.0004, 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(0.0007, 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(0.0009, 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(0.0008, 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(0.0044, 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(0.0009, 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(0.0114, 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(0.0014, 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(0.0014, 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(0.0003, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6278e-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(0.0028, 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(0.0002, 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(0.0011, 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(0.0016, 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(0.0025, 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(0.0015, 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(0.0004, 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(0.0025, 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(6.0373e-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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8940e-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(0.0023, 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(0.0030, 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(5.5559e-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(0.0010, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2160e-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(0.0013, 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(0.0010, 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(0.0001, 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(0.0008, 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(0.0030, 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(8.7014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8307e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0889e-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(0.0014, 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(0.0012, 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(0.0049, 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(0.0012, 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(0.0078, 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(0.0023, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4190e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8941e-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(0.0009, 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(0.0007, 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(9.1041e-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(0.0022, 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(0.0025, 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(0.0019, 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(5.9677e-05, 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(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(0.0092, device='cuda:0', grad_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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7741e-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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0054, 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(7.3694e-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(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(0.0034, 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(0.0017, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8306e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-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(0.0016, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4148e-05, 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(0.0024, 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(0.0012, 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(0.0028, 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(0.0033, 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(0.0031, 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(0.0008, 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(7.6600e-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(0.0086, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-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(0.0019, 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(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7963e-05, 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(0.0031, 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(0.0010, 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(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-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(0.0007, 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(0.0039, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9395e-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(0.0029, 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(9.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, 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(7.2543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5591e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-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(0.0007, 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(4.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6651e-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(0.0065, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2141e-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(0.0001, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8486e-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(0.0001, 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(0.0018, 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(0.0002, 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(0.0040, device='cuda:0', grad_fn=<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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3602e-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(0.0026, 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(0.0008, 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(0.0007, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-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(0.0018, 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(6.4823e-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(8.2292e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-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(0.0018, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-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(0.0002, 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(0.0076, 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(0.0014, 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(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(0.0013, 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(0.0029, device='cuda:0', grad_fn=<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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9064e-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(0.0020, 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(0.0022, 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(0.0017, 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(0.0070, 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(0.0005, 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(0.0008, 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(0.0004, 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(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6854e-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(9.9546e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-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(0.0022, 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(0.0011, 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(0.0034, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2988e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5984e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8184e-05, 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(0.0025, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3140e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1000e-05, 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(5.1956e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0025, 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(0.0022, 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(0.0034, 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(0.0025, 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(0.0011, 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(0.0015, 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(0.0015, 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(0.0004, 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(0.0040, 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(0.0014, 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(0.0023, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7882e-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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8393782383419689, 'agreement': 0.6076923076923076, 'direct_attack': 0.36507936507936506, 'undercutter_attack': 0.32459016393442625, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9418604651162791, 'agreement': 0.28214285714285714, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.6971830985915493, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8876712328767123, 'agreement': 0.3853658536585366, 'direct_attack': 0.2169811320754717, 'undercutter_attack': 0.4429530201342282, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6904205607476636, 'recall': 0.6904205607476636, 'f1': 0.6904205607476636, 'support': None}, 'macro_avg': {'precision': 0.4594908721524707, 'recall': 0.4316235288595035, 'f1': 0.40841242956717155, 'support': None}, 'weighted_avg': {'precision': 0.6742982908534632, 'recall': 0.6904205607476636, 'f1': 0.6606890058308249, 'support': None}}
Loss: tensor(0.0007, 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(7.8598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-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(0.0018, 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(0.0002, 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(8.4441e-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(0.0019, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0007, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-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(0.0073, 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(0.0006, 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(0.0009, 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(0.0003, 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(8.3957e-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(0.0013, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5005e-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(6.0040e-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(8.1988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-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(5.7073e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0003, 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(0.0001, 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(5.5953e-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(7.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0001, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8950e-05, 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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9639e-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(0.0022, 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(0.0012, 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(0.0001, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4462e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4420e-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(0.0006, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9861e-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(0.0001, 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(0.0026, 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(0.0009, 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(4.2631e-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(0.0001, 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(0.0018, 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(0.0010, 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(0.0001, 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(0.0028, 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(0.0002, 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(0.0032, 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(0.0022, 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(0.0001, 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(0.0010, 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(0.0002, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8720e-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(9.4795e-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(0.0001, 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(0.0011, 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(8.5440e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7295e-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(0.0023, 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(0.0009, 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(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8013e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7811e-05, 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(0.0019, 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(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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-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(0.0016, 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(8.9343e-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(0.0015, 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(0.0017, 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(9.9911e-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(0.0012, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-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(6.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0475e-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(0.0007, 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(0.0001, 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(0.0021, 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(0.0010, 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(0.0003, 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(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(0.0005, 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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0797e-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(9.3887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-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(0.0019, 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(0.0032, 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(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-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(0.0009, 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(0.0007, 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(0.0040, 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(0.0019, 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(0.0003, 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(0.0028, 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(0.0059, 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(0.0011, 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(0.0006, 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(6.4066e-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(9.8003e-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(0.0027, 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(0.0005, 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(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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0584e-05, 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(0.0006, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5298e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-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(0.0011, 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(0.0021, 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(0.0019, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5934e-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(9.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5620e-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(0.0009, 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(0.0025, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-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(0.0015, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-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(0.0049, 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(0.0012, 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(0.0010, 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(0.0043, 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(0.0004, 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(7.6722e-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(0.0016, device='cuda:0', grad_fn=<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(0.0021, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7096e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-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(0.0013, 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(0.0011, 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(0.0003, 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(6.4218e-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(0.0001, 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(0.0003, 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(0.0006, 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(0.0005, 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(0.0013, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8084e-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(0.0051, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6752e-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(6.6095e-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(0.0020, 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(0.0021, 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(0.0010, 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(8.3049e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0424e-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(0.0008, 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(0.0009, 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(0.0027, 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(0.0016, 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(0.0006, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-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(0.0020, 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(0.0029, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1908e-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(0.0027, 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(0.0001, 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(0.0024, 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(0.0019, 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(0.0012, 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(9.8366e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6085e-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(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(0.0002, 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(0.0006, 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(5.9313e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8669e-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(0.0034, device='cuda:0', grad_fn=<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(0.0008, 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(0.0025, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(9.6520e-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(0.0003, 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(0.0019, 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(0.0054, 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(0.0018, 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(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-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(0.0011, 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(4.7990e-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(0.0003, 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(7.7600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-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(0.0001, 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(0.0019, 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(0.0017, 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(0.0015, 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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8402417962003454, 'agreement': 0.6046511627906976, 'direct_attack': 0.36666666666666664, 'undercutter_attack': 0.3246753246753247, 'partial': 0.15789473684210525}, 'recall': {'support': 0.9428294573643411, 'agreement': 0.2785714285714286, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8885844748858448, 'agreement': 0.38141809290953543, 'direct_attack': 0.21052631578947367, 'undercutter_attack': 0.4444444444444445, 'partial': 0.10843373493975904}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6904205607476636, 'recall': 0.6904205607476636, 'f1': 0.6904205607476636, 'support': None}, 'macro_avg': {'precision': 0.458825937435028, 'recall': 0.43116921041986095, 'f1': 0.4066814125938115, 'support': None}, 'weighted_avg': {'precision': 0.6742871583020859, 'recall': 0.6904205607476636, 'f1': 0.6601139330426625, 'support': None}}
Loss: tensor(5.8677e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, 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(0.0006, 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(0.0020, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2251e-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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-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(0.0016, 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(7.6236e-05, 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(0.0023, 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(6.0464e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-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(0.0005, 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(0.0008, 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(7.4299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-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(0.0008, 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(7.6752e-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(0.0021, 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(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1150e-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(0.0001, 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(0.0004, 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(0.0007, 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(0.0011, 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(0.0014, 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(0.0023, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6470e-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(8.3594e-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(0.0033, 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(0.0033, 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(0.0018, 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(8.3140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-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(4.8414e-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(9.5187e-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(0.0028, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0796e-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(9.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5972e-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(6.0433e-05, 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(0.0001, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9204e-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(0.0007, 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(0.0017, 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(0.0019, 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(0.0013, 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(0.0022, 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(0.0058, 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(0.0001, 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(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(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(0.0012, 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(8.6740e-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(5.8980e-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(0.0003, 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(0.0015, 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(5.8284e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-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(0.0008, 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(0.0006, 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(0.0010, 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(0.0011, 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(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-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(6.7124e-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(0.0021, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-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(0.0001, 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(7.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0012, 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(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(6.9365e-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(0.0007, 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(0.0016, 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(7.4269e-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(0.0016, 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(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-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(0.0114, 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(0.0004, 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(0.0005, 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(8.3260e-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(0.0005, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, 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(0.0022, 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(0.0001, 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(8.9830e-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(0.0001, 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(0.0001, 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(6.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-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(0.0017, 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(0.0001, 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(0.0006, 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(9.5400e-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(0.0014, 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(0.0013, device='cuda:0', grad_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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6488e-05, device='cuda:0', grad_fn=<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(5.6377e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-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(9.0949e-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(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(0.0020, 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(0.0024, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-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(8.3775e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-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(0.0004, 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(0.0004, 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(0.0007, 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(0.0019, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0012, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, 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(0.0017, 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(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3745e-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(8.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1212e-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(0.0016, 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(0.0008, 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(7.6782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-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(0.0014, 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(0.0018, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-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(0.0025, 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(0.0021, 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(0.0005, 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(6.7457e-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(0.0007, 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(0.0008, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6712e-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(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(8.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0016, 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(0.0013, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3057e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-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(0.0086, device='cuda:0', grad_fn=<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(0.0006, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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(0.0007, 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(0.0011, 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(0.0007, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5135e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, 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(7.0939e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-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(8.9071e-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(0.0002, 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(0.0005, 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(0.0024, 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(7.9536e-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(0.0008, 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(0.0001, 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(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6319e-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(0.0007, 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(0.0005, 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(4.0572e-05, device='cuda:0', grad_fn=<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(8.9102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-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(0.0039, 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(0.0040, 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(9.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2199e-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(0.0002, 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(0.0017, 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(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(7.5843e-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(8.3442e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-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(7.2149e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8718e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399e-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(0.0018, 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(0.0003, 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(0.0005, 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(0.0015, 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(0.0039, 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(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8396551724137931, 'agreement': 0.6015625, 'direct_attack': 0.36666666666666664, 'undercutter_attack': 0.3257328990228013, 'partial': 0.15789473684210525}, 'recall': {'support': 0.9437984496124031, 'agreement': 0.275, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8886861313868614, 'agreement': 0.37745098039215685, 'direct_attack': 0.21052631578947367, 'undercutter_attack': 0.44543429844098, 'partial': 0.10843373493975904}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6904205607476636, 'recall': 0.6904205607476636, 'f1': 0.6904205607476636, 'support': None}, 'macro_avg': {'precision': 0.4583023949890733, 'recall': 0.4306487231551877, 'f1': 0.4061062921898461, 'support': None}, 'weighted_avg': {'precision': 0.6735161035288523, 'recall': 0.6904205607476636, 'f1': 0.6596084875237904, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, 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(0.0010, 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(0.0003, 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(0.0071, device='cuda:0', grad_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(0.0003, 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(8.7347e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4510e-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(6.0222e-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(0.0002, 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(5.5044e-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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-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(0.0001, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6702e-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(0.0008, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7246e-05, 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(0.0014, 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(0.0001, 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(0.0024, 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(7.5390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6348e-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(6.9939e-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(0.0011, 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(0.0008, 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(0.0026, 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(6.0887e-05, 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(0.0017, 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(0.0012, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-05, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-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(0.0004, 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(4.1450e-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(0.0005, 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(5.1835e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-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(9.2008e-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(9.0222e-05, 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(0.0017, 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(8.1747e-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(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1836e-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(8.4260e-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(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1021e-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(0.0009, 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(0.0005, 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(8.4712e-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(0.0025, device='cuda:0', grad_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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4560e-05, device='cuda:0', grad_fn=<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(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(0.0003, 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(0.0013, 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(0.0003, 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(0.0006, 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(0.0024, 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(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-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(0.0018, 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(7.5117e-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(6.4460e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5821e-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(0.0017, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(6.4793e-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(0.0013, 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(5.8435e-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(0.0014, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-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(0.0009, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, 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(0.0016, 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(0.0012, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5773e-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(0.0004, 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(0.0002, 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(0.0011, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1533e-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(4.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-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(5.7103e-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(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2522e-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(0.0009, 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(9.0284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, 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(0.0011, 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(5.5862e-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(6.1433e-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(5.8799e-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(0.0021, 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(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(0.0017, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(5.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-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(0.0006, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-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(0.0015, 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(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1938e-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(5.6316e-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(0.0008, 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(0.0017, 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(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, 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(7.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7015e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-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(0.0006, 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(0.0018, 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(0.0006, 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(0.0016, 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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7831e-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(5.3379e-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(0.0012, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6880e-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(4.6718e-05, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8968e-05, device='cuda:0', grad_fn=<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(7.8387e-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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-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(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(0.0022, 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(0.0019, 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(9.1038e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-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(6.9272e-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(0.0029, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4289e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0019, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(9.1737e-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(5.9343e-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(0.0004, 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(0.0001, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(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(4.7233e-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(4.3328e-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(0.0027, 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(0.0018, 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(5.0836e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7660e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2582e-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(0.0007, 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(0.0027, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4066e-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(6.7154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-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(0.0023, 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(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9042e-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(0.0010, 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(0.0016, 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(6.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3572e-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(4.2601e-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(5.8829e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-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(0.0017, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1978e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7488e-05, 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(0.0020, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6115e-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(0.0005, 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(0.0011, 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(0.0003, 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(0.0001, 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(5.1442e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7174e-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(0.0007, 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(9.2463e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-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(8.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0989e-05, 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(0.0004, 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(8.6346e-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(0.0010, 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(9.3675e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-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(0.0007, 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(7.1757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8409286328460877, 'agreement': 0.6062992125984252, 'direct_attack': 0.38333333333333336, 'undercutter_attack': 0.32679738562091504, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9467570183930301, 'agreement': 0.27598566308243727, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8907103825136612, 'agreement': 0.37931034482758613, 'direct_attack': 0.2200956937799043, 'undercutter_attack': 0.44642857142857145, 'partial': 0.1090909090909091}, 'support': {'support': 1033, 'agreement': 279, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6933411214953271, 'recall': 0.6933411214953271, 'f1': 0.6933411214953271, 'support': None}, 'macro_avg': {'precision': 0.4636145700226094, 'recall': 0.43277985140699515, 'f1': 0.40912718032812645, 'support': None}, 'weighted_avg': {'precision': 0.6769138496569291, 'recall': 0.6933411214953271, 'f1': 0.6623886892234117, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, 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(0.0003, 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(0.0014, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0334e-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(0.0018, 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(0.0004, 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(3.4002e-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(8.0959e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9426e-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(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-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(7.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-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(0.0011, 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(0.0004, 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(9.5096e-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(0.0026, 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(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, 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(0.0003, 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(0.0007, 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(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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-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(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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2967e-05, device='cuda:0', grad_fn=<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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4673e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5097e-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(0.0024, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-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(0.0025, 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(0.0016, 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(0.0026, 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(6.8243e-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(0.0009, 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(6.8274e-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(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-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(0.0039, 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(7.4088e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-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(0.0007, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6145e-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(0.0003, 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(0.0042, 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(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-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(5.9797e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8973e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-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(5.1835e-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(0.0006, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5298e-05, 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(0.0008, 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(5.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4916e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0006, 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(0.0007, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-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(0.0021, 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(0.0023, 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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-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(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(4.2995e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5983e-05, 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(0.0006, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-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(0.0015, 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(7.9899e-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(0.0004, 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(8.8043e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-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(0.0007, 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(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5832e-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(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(0.0008, 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(0.0005, 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(0.0010, 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(0.0017, 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(0.0016, 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(0.0002, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9677e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-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(7.7327e-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(6.8124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(0.0007, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0005, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0939e-05, 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(0.0011, 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(9.5398e-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(0.0003, 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(6.7972e-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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-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(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8949e-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(7.9809e-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(0.0008, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-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(3.7545e-05, 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(0.0014, 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(4.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-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(8.1534e-05, 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(5.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-05, 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(0.0006, 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(9.8364e-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(0.0028, 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(0.0012, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-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(0.0012, 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(0.0002, 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(0.0006, 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(0.0028, 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(4.0966e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0202e-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(0.0034, 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(0.0004, 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(6.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3430e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0282e-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(7.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-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(7.3330e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1907e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-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(0.0012, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-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(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9121e-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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<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(5.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4593e-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(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(4.1632e-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(0.0012, 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(0.0003, 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(6.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2199e-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(0.0001, 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(0.0014, 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(0.0024, 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(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8276e-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(9.7668e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-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(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9374e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8517e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-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(0.0004, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-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(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-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(5.1048e-05, 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(0.0002, 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(6.4307e-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(4.4418e-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(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-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(0.0025, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8400687876182287, 'agreement': 0.6062992125984252, 'direct_attack': 0.3728813559322034, 'undercutter_attack': 0.3257328990228013, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9467054263565892, 'agreement': 0.275, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8902050113895217, 'agreement': 0.3783783783783784, 'direct_attack': 0.21153846153846156, 'undercutter_attack': 0.44543429844098, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6921728971962616, 'recall': 0.6921728971962616, 'f1': 0.6921728971962616, 'support': None}, 'macro_avg': {'precision': 0.46113930817718884, 'recall': 0.4312301185040249, 'f1': 0.4069294117676502, 'support': None}, 'weighted_avg': {'precision': 0.6752605252263809, 'recall': 0.6921728971962616, 'f1': 0.6608056822071796, 'support': None}}
Loss: tensor(6.1613e-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(0.0095, 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(0.0017, 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(0.0014, 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(0.0014, 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(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-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(0.0008, 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(0.0009, 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(6.5791e-05, device='cuda:0', grad_fn=<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(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8393e-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(0.0015, 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(0.0001, 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(0.0003, 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(4.6295e-05, device='cuda:0', grad_fn=<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(3.8907e-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(0.0011, 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(5.7345e-05, 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(0.0010, 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(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-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(0.0004, 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(0.0004, 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(6.3522e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5419e-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(0.0002, 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(3.2700e-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(0.0004, 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(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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3248e-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(6.7094e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2915e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4169e-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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-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(0.0001, 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(9.4793e-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(7.4602e-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(0.0006, 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(7.7084e-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(0.0010, 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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-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(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5855e-05, 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(7.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2313e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-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(0.0003, 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(0.0005, 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(0.0010, 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(0.0012, 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(0.0012, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8241e-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(9.5340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-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(4.8142e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4389e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0018, 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(0.0001, 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(0.0011, 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(0.0008, 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(0.0008, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7276e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-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(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-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(4.6839e-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(5.3682e-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(0.0022, 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(0.0021, 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(7.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-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(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5852e-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(8.1898e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7223e-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(0.0007, 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(5.5286e-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(8.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-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(0.0005, 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(0.0002, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8101e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-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(3.5092e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2291e-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(8.3138e-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(3.2943e-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(0.0006, 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(0.0015, 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(6.4610e-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(4.8687e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5278e-05, 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(0.0003, 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(0.0011, 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(7.1847e-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(0.0015, 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(0.0019, 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(0.0006, 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(0.0002, 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(6.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5734e-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(6.7819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-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(6.8395e-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(0.0015, 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(0.0004, 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(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-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(4.5084e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0202e-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(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0016, 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(0.0009, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2948e-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(3.4245e-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(7.4693e-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(8.3200e-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(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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3036e-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(8.0657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2270e-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(0.0012, 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(0.0008, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3542e-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(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(0.0002, 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(0.0006, 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(4.6839e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1333e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, 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(0.0007, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4753e-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(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0311e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-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(3.9937e-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(0.0013, 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(0.0005, 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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-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(5.4196e-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(0.0004, 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(0.0001, 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(0.0011, 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(4.9716e-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(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-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(0.0021, 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(0.0003, 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(0.0001, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8122e-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(5.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7942e-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(0.0009, 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(2.9642e-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(0.0008, 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(0.0004, 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(0.0002, 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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-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(0.0009, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-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(0.0010, 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(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-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(0.0017, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8400687876182287, 'agreement': 0.6062992125984252, 'direct_attack': 0.3728813559322034, 'undercutter_attack': 0.3257328990228013, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9467054263565892, 'agreement': 0.275, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8902050113895217, 'agreement': 0.3783783783783784, 'direct_attack': 0.21153846153846156, 'undercutter_attack': 0.44543429844098, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6921728971962616, 'recall': 0.6921728971962616, 'f1': 0.6921728971962616, 'support': None}, 'macro_avg': {'precision': 0.46113930817718884, 'recall': 0.4312301185040249, 'f1': 0.4069294117676502, 'support': None}, 'weighted_avg': {'precision': 0.6752605252263809, 'recall': 0.6921728971962616, 'f1': 0.6608056822071796, 'support': None}}
Loss: tensor(6.5942e-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(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6730e-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(7.7447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-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(0.0003, 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(5.1896e-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(4.7203e-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(0.0015, 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(8.5379e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-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(0.0025, 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(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(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(0.0011, 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(0.0007, 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(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-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(0.0003, 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(0.0009, 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(5.2410e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1068e-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(3.9906e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-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(9.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-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(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(0.0002, 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(0.0004, 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(8.9527e-05, 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(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-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(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-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(0.0012, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, 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(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-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(0.0010, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5510e-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(4.1934e-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(3.2095e-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(5.0624e-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(0.0005, 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(0.0004, 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(0.0002, 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(7.1604e-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(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-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(6.4097e-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(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2786e-05, device='cuda:0', grad_fn=<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(4.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-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(0.0019, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(2.4707e-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(0.0018, 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(0.0005, 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(0.0003, 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(0.0001, 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(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(0.0002, 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(7.8326e-05, device='cuda:0', grad_fn=<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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0797e-05, 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(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6185e-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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-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(0.0009, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9941e-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(0.0020, 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(0.0008, 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(0.0004, 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(0.0010, 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(0.0010, 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(0.0009, 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(0.0020, 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(0.0002, 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(0.0003, 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(0.0013, 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(0.0001, 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(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5781e-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(3.8786e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-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(8.9221e-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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-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(0.0018, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-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(0.0010, 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(0.0010, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-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(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(0.0013, 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(3.4335e-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(0.0015, 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(0.0005, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8040e-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(0.0015, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-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(0.0004, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6580e-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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0020, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-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(7.1454e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6909e-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(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-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(9.1554e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-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(0.0015, 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(0.0010, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-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(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(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5671e-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(5.0685e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4458e-05, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0001, 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(4.6446e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, 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(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(0.0016, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6879e-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(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(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-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(0.0001, 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(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-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(0.0014, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, 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(5.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, 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(3.8665e-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(0.0007, 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(0.0007, 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(0.0013, 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(0.0008, 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(0.0022, 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(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8400687876182287, 'agreement': 0.6062992125984252, 'direct_attack': 0.3728813559322034, 'undercutter_attack': 0.3257328990228013, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9467054263565892, 'agreement': 0.275, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8902050113895217, 'agreement': 0.3783783783783784, 'direct_attack': 0.21153846153846156, 'undercutter_attack': 0.44543429844098, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6921728971962616, 'recall': 0.6921728971962616, 'f1': 0.6921728971962616, 'support': None}, 'macro_avg': {'precision': 0.46113930817718884, 'recall': 0.4312301185040249, 'f1': 0.4069294117676502, 'support': None}, 'weighted_avg': {'precision': 0.6752605252263809, 'recall': 0.6921728971962616, 'f1': 0.6608056822071796, 'support': None}}
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(0.0007, 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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4862e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-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(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7747e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-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(9.4369e-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(0.0006, 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(0.0010, 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(0.0018, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-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(3.0642e-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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(0.0004, 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(5.4893e-05, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-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(9.3611e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(0.0008, 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(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(0.0003, 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(0.0004, 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(8.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, 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(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-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(4.0330e-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(0.0002, 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(0.0006, 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(0.0004, 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(0.0003, 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(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4892e-05, 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(0.0010, 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(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(0.0004, 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(0.0008, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-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(4.7960e-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(0.0012, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6074e-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(0.0002, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-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(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-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(0.0009, 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(0.0005, 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(0.0005, 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(7.2634e-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(3.2398e-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(5.5226e-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(0.0010, 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(0.0006, 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(4.6597e-05, 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(0.0004, 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(7.3995e-05, 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(0.0001, 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(0.0007, 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(0.0012, 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(0.0004, 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(8.4714e-05, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4823e-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(4.6295e-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(0.0007, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<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(3.9028e-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(0.0003, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-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(5.7951e-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(5.6254e-05, 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(0.0006, 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(5.0291e-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(8.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-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(0.0001, 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(0.0006, 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(3.8120e-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(6.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8204e-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(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5276e-05, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-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(3.2731e-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(0.0005, 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(6.3550e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, 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(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(4.4024e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9275e-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(4.2086e-05, 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(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-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(6.5851e-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(5.4045e-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(0.0014, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6951e-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(0.0005, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-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(0.0007, 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(3.0975e-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(4.3630e-05, 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(4.9141e-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(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2685e-05, device='cuda:0', grad_fn=<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(5.0594e-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(0.0025, 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(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(7.1181e-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(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-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(0.0064, device='cuda:0', grad_fn=<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(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-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(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1987e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-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(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-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(4.5235e-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(0.0016, 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(3.3366e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, 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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-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(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-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(5.8677e-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(0.0020, 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(0.0016, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-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(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(5.9707e-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(3.9240e-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(6.5550e-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(0.0010, 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(2.6675e-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(2.6675e-05, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6449e-05, device='cuda:0', grad_fn=<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(0.0002, 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(7.0182e-05, 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(0.0008, 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(2.9582e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0005, 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(0.0009, 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(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-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(0.0007, 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(0.0013, 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(0.0003, 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(6.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1262e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8407917383820999, 'agreement': 0.609375, 'direct_attack': 0.3728813559322034, 'undercutter_attack': 0.3257328990228013, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9467054263565892, 'agreement': 0.2785714285714286, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8906107566089335, 'agreement': 0.38235294117647056, 'direct_attack': 0.21153846153846156, 'undercutter_attack': 0.44543429844098, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6927570093457944, 'recall': 0.6927570093457944, 'f1': 0.6927570093457944, 'support': None}, 'macro_avg': {'precision': 0.461899055810278, 'recall': 0.4319444042183106, 'f1': 0.4078054733711509, 'support': None}, 'weighted_avg': {'precision': 0.6761993719908412, 'recall': 0.6927570093457944, 'f1': 0.6617003122596906, 'support': None}}
Loss: tensor(4.4478e-05, 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(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-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(3.7545e-05, device='cuda:0', grad_fn=<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(4.0845e-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(3.2034e-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(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1433e-05, 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(8.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, 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(0.0001, 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(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-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(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-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(0.0010, 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(0.0001, 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(8.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-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(0.0003, 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(0.0001, 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(0.0015, 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(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-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(6.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7195e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0003, 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(0.0001, 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(0.0011, 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(7.8235e-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(0.0013, 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(0.0003, 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(0.0004, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5885e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-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(2.7099e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9311e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0012, 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(0.0005, 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(0.0006, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2289e-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(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(8.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6974e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0018, 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(0.0008, 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(0.0001, 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(4.3630e-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(4.6476e-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(4.5084e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, 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(2.4919e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0005, 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(8.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(6.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-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(0.0005, 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(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(2.9188e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-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(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-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(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8626e-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(0.0004, 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(0.0001, 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(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0513e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8337e-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(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2885e-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(5.5195e-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(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3160e-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(0.0003, 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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8959e-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(2.9975e-05, 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(5.2138e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-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(6.0099e-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(4.4054e-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(7.9292e-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(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(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4963e-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(0.0005, 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(5.8586e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, 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(0.0012, 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(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-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(7.6934e-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(4.7536e-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(0.0004, 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(2.1891e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(6.8696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, 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(4.8172e-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(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-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(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(2.2618e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-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(0.0001, 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(9.5035e-05, 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(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(4.0633e-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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0006, 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(0.0001, 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(3.1822e-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(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(0.0016, 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(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(7.0182e-05, 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(0.0014, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0142e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-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(3.0248e-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(8.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-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(0.0003, 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(0.0004, 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(0.0014, 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(0.0003, 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(4.9473e-05, 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(6.9789e-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(0.0014, 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(5.5468e-05, device='cuda:0', grad_fn=<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(3.1096e-05, 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(0.0009, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0756e-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(3.5819e-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(7.2906e-05, device='cuda:0', grad_fn=<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(0.0003, 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(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-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(0.0001, 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(7.3480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-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(3.4396e-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(4.0724e-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(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(0.0007, 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(6.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-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(4.5780e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0685e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5085e-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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8409286328460877, 'agreement': 0.609375, 'direct_attack': 0.3728813559322034, 'undercutter_attack': 0.32679738562091504, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9476744186046512, 'agreement': 0.2785714285714286, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8911161731207289, 'agreement': 0.38235294117647056, 'direct_attack': 0.21153846153846156, 'undercutter_attack': 0.44642857142857145, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6933411214953271, 'recall': 0.6933411214953271, 'f1': 0.6933411214953271, 'support': None}, 'macro_avg': {'precision': 0.4621393320226984, 'recall': 0.4321382026679229, 'f1': 0.4081054112710283, 'support': None}, 'weighted_avg': {'precision': 0.6763701851822942, 'recall': 0.6933411214953271, 'f1': 0.6620874481267529, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-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(6.2521e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9623e-05, device='cuda:0', grad_fn=<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(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-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(0.0002, 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(0.0004, 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(0.0007, 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(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4188e-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(3.6424e-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(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, 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(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-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(5.8707e-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(0.0005, 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(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-05, device='cuda:0', grad_fn=<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(2.5979e-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(2.5404e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7577e-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(9.5945e-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(0.0008, 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(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8767e-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(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(6.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4792e-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(0.0008, 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(3.6909e-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(3.6424e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-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(0.0004, 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(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0005, 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(6.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7476e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0004, 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(0.0004, 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(0.0005, 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(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7770e-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(4.5598e-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(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0007, 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(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-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(3.0248e-05, 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(0.0008, 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(0.0006, 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(7.0664e-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(0.0007, 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(2.5161e-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(0.0009, 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(0.0007, 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(0.0030, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-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(7.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(0.0005, 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(9.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-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(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-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(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6599e-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(8.9857e-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(6.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0001e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5146e-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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-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(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-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(0.0008, 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(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-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(9.9028e-05, 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(9.4794e-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(6.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(9.1308e-05, 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(3.4426e-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(8.9977e-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(0.0018, 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(0.0001, 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(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-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(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-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(4.1753e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5077e-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(0.0003, 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(0.0009, 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(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6042e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-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(0.0004, 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(3.7454e-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(0.0008, 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(2.6766e-05, 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(0.0003, 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(4.6203e-05, device='cuda:0', grad_fn=<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(0.0004, 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(6.0553e-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(2.5434e-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(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-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(4.5628e-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(0.0001, 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(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-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(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, 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(7.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-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(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(9.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0019, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4590e-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(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-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(0.0015, 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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-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(0.0004, 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(0.0003, 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(4.1541e-05, 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(0.0006, 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(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(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1491e-05, 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(0.0005, 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(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, 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(9.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3834e-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(0.0003, 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(9.2960e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0004, 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(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6306e-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(0.0002, 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(2.8946e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-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(0.0004, 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(3.3851e-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(9.4642e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6327e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-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(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-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(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0002, 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(3.2397e-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(0.0001, 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(0.0005, 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(0.0008, 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(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(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-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(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(6.1855e-05, device='cuda:0', grad_fn=<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(7.4449e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1674e-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(0.0019, 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(2.6433e-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(8.2836e-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(0.0005, 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(6.3552e-05, device='cuda:0', grad_fn=<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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8416523235800344, 'agreement': 0.609375, 'direct_attack': 0.36065573770491804, 'undercutter_attack': 0.32786885245901637, 'partial': 0.16071428571428573}, 'recall': {'support': 0.9476744186046512, 'agreement': 0.2785714285714286, 'direct_attack': 0.1476510067114094, 'undercutter_attack': 0.704225352112676, 'partial': 0.08256880733944955}, 'f1': {'support': 0.8915223336371922, 'agreement': 0.38235294117647056, 'direct_attack': 0.20952380952380956, 'undercutter_attack': 0.447427293064877, 'partial': 0.1090909090909091}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6933411214953271, 'recall': 0.6933411214953271, 'f1': 0.6933411214953271, 'support': None}, 'macro_avg': {'precision': 0.4600532398916509, 'recall': 0.4321382026679229, 'f1': 0.4079834572986517, 'support': None}, 'weighted_avg': {'precision': 0.6758312716382393, 'recall': 0.6933411214953271, 'f1': 0.6622397804720581, 'support': None}}
Loss: tensor(2.7644e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0007, 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(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-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(3.6152e-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(0.0001, device='cuda:0', grad_fn=<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(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-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(1.9499e-05, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8710e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3240e-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(0.0005, 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(9.0703e-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(8.6316e-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(1.6805e-05, device='cuda:0', grad_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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3126e-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(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-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(2.4465e-05, 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(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(0.0004, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-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(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, 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(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(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(9.1432e-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(5.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-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(0.0001, 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(2.6100e-05, device='cuda:0', grad_fn=<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(4.0209e-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(9.5520e-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(3.8816e-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(2.8795e-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(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, 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(5.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0006, 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(2.8552e-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(5.1139e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2391e-05, device='cuda:0', grad_fn=<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(5.3863e-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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7941e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, 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(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(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<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(2.7614e-05, device='cuda:0', grad_fn=<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(8.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9566e-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(4.1269e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, 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(0.0002, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-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(4.5871e-05, device='cuda:0', grad_fn=<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(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(2.0014e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-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(4.2177e-05, 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(3.7848e-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(3.8483e-05, 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(7.6479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9679e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3399e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5588e-05, 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(0.0002, 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(3.7938e-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(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4619e-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(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(2.8219e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, 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(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-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(0.0002, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8425e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2219e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0414e-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(3.5153e-05, 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(0.0002, 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(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, 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(0.0011, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-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(0.0002, 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(8.0829e-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(5.6952e-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(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(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0006, 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(4.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-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(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(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-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(5.5165e-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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5418e-05, device='cuda:0', grad_fn=<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(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6466e-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(9.9936e-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(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(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(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(7.2269e-05, device='cuda:0', grad_fn=<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(9.3096e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<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(0.0002, 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(4.4569e-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(3.9119e-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(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0564e-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(7.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-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(0.0002, 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(0.0009, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, 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(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, 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(0.0005, 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(5.9645e-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(3.0611e-05, device='cuda:0', grad_fn=<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(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, 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(0.0005, 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(3.5728e-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(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-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(0.0006, 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(0.0005, 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(0.0009, 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(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(3.0127e-05, device='cuda:0', grad_fn=<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(0.0002, 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(3.4880e-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(0.0006, 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(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, 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(1.7804e-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(8.5015e-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(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(2.6282e-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(0.0007, 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(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-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(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-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(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7246e-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(0.0002, 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(0.0002, 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(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8409286328460877, 'agreement': 0.6062992125984252, 'direct_attack': 0.3770491803278688, 'undercutter_attack': 0.32679738562091504, 'partial': 0.14545454545454545}, 'recall': {'support': 0.9476744186046512, 'agreement': 0.275, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.07339449541284404}, 'f1': {'support': 0.8911161731207289, 'agreement': 0.3783783783783784, 'direct_attack': 0.21904761904761905, 'undercutter_attack': 0.44642857142857145, 'partial': 0.0975609756097561}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6927570093457944, 'recall': 0.6927570093457944, 'f1': 0.6927570093457944, 'support': None}, 'macro_avg': {'precision': 0.45930579136956845, 'recall': 0.43093133644751075, 'f1': 0.40650634351701076, 'support': None}, 'weighted_avg': {'precision': 0.6752583123284399, 'recall': 0.6927570093457944, 'f1': 0.6613568547482209, 'support': None}}
Loss: tensor(8.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4400e-05, device='cuda:0', grad_fn=<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(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(8.3923e-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(5.7586e-05, 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(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-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(6.5245e-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(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(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2751e-05, device='cuda:0', grad_fn=<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(0.0008, 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(2.6887e-05, 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(1.6199e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5316e-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(2.7372e-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(8.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-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(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-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(1.9651e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-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(2.4768e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(6.6063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1659e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9839e-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(2.2921e-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(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-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(0.0003, 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(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-05, 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(0.0005, 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(8.5924e-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(0.0006, 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(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4660e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5852e-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(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-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(0.0007, 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(2.3647e-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(2.2376e-05, 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(0.0004, 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(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4044e-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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5963e-05, device='cuda:0', grad_fn=<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(6.3610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-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(5.3015e-05, device='cuda:0', grad_fn=<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(4.0451e-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(0.0003, 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(0.0002, 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(4.2843e-05, device='cuda:0', grad_fn=<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(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9827e-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(0.0002, 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(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3822e-05, 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(6.6456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1119e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3025e-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(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-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(4.2450e-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(3.3336e-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(9.0866e-05, device='cuda:0', grad_fn=<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(1.8379e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3340e-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(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0003, 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(6.8365e-05, device='cuda:0', grad_fn=<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(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-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(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5509e-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(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(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(0.0006, 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(5.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-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(0.0002, 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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1887e-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(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(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-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(2.1922e-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(0.0008, 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(2.5192e-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(1.9530e-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(5.3773e-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(7.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9217e-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(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9340e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-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(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9949e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-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(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-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(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1470e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0368e-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(2.7462e-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(0.0001, 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(0.0001, 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(6.6125e-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(4.3630e-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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-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(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6041e-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(6.6153e-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(4.2359e-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(0.0002, 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(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(6.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-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(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0643e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4569e-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(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-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(0.0005, 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(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(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7869e-05, device='cuda:0', grad_fn=<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(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1343e-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(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0004, 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(2.7281e-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(8.2604e-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(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(2.5131e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<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(3.2731e-05, 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(2.3012e-05, device='cuda:0', grad_fn=<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(8.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5166e-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(5.7102e-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(0.0003, 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(9.6612e-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(2.9218e-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(2.6403e-05, 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(0.0007, 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(0.0002, 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(3.7877e-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(0.0003, 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(2.4798e-05, 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(0.0004, 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(3.6425e-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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.4314e-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(0.0010, 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(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(3.3185e-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(0.0001, 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(0.0001, 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(4.1510e-05, 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(0.0004, 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(2.3375e-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(2.6372e-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(0.0005, 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(3.7696e-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(2.7523e-05, 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(0.0002, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8410652920962199, 'agreement': 0.6062992125984252, 'direct_attack': 0.3770491803278688, 'undercutter_attack': 0.32679738562091504, 'partial': 0.14814814814814814}, 'recall': {'support': 0.9477250726040658, 'agreement': 0.275, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.07407407407407407}, 'f1': {'support': 0.8912152935821576, 'agreement': 0.3783783783783784, 'direct_attack': 0.21904761904761905, 'undercutter_attack': 0.44642857142857145, 'partial': 0.09876543209876543}, 'support': {'support': 1033, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 108}, 'micro_avg': {'precision': 0.6933411214953271, 'recall': 0.6933411214953271, 'f1': 0.6933411214953271, 'support': None}, 'macro_avg': {'precision': 0.45987184375831536, 'recall': 0.4310773829796397, 'f1': 0.40676705890709836, 'support': None}, 'weighted_avg': {'precision': 0.6759169292581638, 'recall': 0.6933411214953271, 'f1': 0.6619561701308025, 'support': None}}
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(3.5728e-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(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(0.0007, 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(0.0002, 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(0.0005, 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(6.8243e-05, 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(0.0002, 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(2.5434e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-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(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(4.1632e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8607e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-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(0.0002, 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(3.2216e-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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-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(8.7902e-05, 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(0.0004, 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(0.0003, 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(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-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(4.4296e-05, 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(7.7416e-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(0.0002, 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(0.0005, 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(7.9202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7921e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<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(2.2345e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-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(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-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(8.0415e-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(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1715e-05, device='cuda:0', grad_fn=<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(8.8965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3565e-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(0.0002, 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(0.0008, 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(0.0006, 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(0.0008, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8157e-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(0.0004, 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(0.0005, 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(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(2.4980e-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(8.8757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-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(0.0002, 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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2670e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-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(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(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(0.0012, 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(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-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(0.0008, 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(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7294e-05, device='cuda:0', grad_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(0.0008, 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(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, 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(0.0001, 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(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, 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(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, 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(0.0004, 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(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-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(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8333e-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(3.4002e-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(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5679e-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(0.0001, 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(0.0003, 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(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2992e-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(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, 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(7.1514e-05, device='cuda:0', grad_fn=<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(3.6394e-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(0.0004, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(7.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9917e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, 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(4.4054e-05, 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(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-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(5.7315e-05, 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(2.6796e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0005, 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(9.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, 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(2.4859e-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(2.6372e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, 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(2.1528e-05, 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(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(0.0004, 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(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, 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(7.6235e-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(0.0002, 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(3.3609e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-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(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9536e-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(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7864e-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(5.4499e-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(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<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(6.2643e-05, 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(6.5518e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0008, 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(0.0001, 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(6.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-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(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0108e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(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(6.1098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4428e-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(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(7.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0789e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, 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(0.0010, 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(1.4473e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3412e-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(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7376e-05, 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(6.6455e-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(0.0003, 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(8.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, 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(9.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1743e-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(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(2.1134e-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(0.0002, 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(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(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-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(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7940e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8454e-05, 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(0.0003, 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(2.4132e-05, 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(0.0009, 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(0.0005, 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(0.0009, 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(0.0004, 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(6.1432e-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(5.2380e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0001, 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(0.0003, 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(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, 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(0.0015, 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(7.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-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(6.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2926e-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(2.3920e-05, 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(0.0005, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8402061855670103, 'agreement': 0.6031746031746031, 'direct_attack': 0.3770491803278688, 'undercutter_attack': 0.3257328990228013, 'partial': 0.14814814814814814}, 'recall': {'support': 0.9476744186046512, 'agreement': 0.2714285714285714, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.07339449541284404}, 'f1': {'support': 0.8907103825136613, 'agreement': 0.37438423645320196, 'direct_attack': 0.21904761904761905, 'undercutter_attack': 0.44543429844098, 'partial': 0.098159509202454}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6921728971962616, 'recall': 0.6921728971962616, 'f1': 0.6921728971962616, 'support': None}, 'macro_avg': {'precision': 0.45886220324808635, 'recall': 0.430217050733225, 'f1': 0.4055472091315832, 'support': None}, 'weighted_avg': {'precision': 0.6743949883599776, 'recall': 0.6921728971962616, 'f1': 0.660414633808865, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-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(0.0004, 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(1.5351e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2039e-05, device='cuda:0', grad_fn=<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(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1967e-05, device='cuda:0', grad_fn=<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(6.6851e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0193e-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(0.0013, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, 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(2.1589e-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(2.0711e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0018, 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(2.7069e-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(0.0004, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<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(1.2202e-05, 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(0.0002, 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(3.3336e-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(5.0140e-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(0.0010, 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(1.9742e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8516e-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(0.0006, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5034e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4097e-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(1.9348e-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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, 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(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7364e-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(2.2527e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-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(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(2.7281e-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(7.9294e-05, 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(0.0008, 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(0.0005, 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(2.8885e-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(6.4882e-05, 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(7.1491e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-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(1.5775e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, 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(5.8071e-05, device='cuda:0', grad_fn=<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(5.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-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(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4541e-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(5.3016e-05, device='cuda:0', grad_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(2.0892e-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>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, 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(0.0002, 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(2.4919e-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(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-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(0.0003, 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(7.4723e-05, 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(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-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(9.4648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9462e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<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(1.9802e-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(1.2475e-05, 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(9.5364e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, 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(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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1101e-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(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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6960e-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(8.4752e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6324e-05, device='cuda:0', grad_fn=<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(0.0007, 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(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(9.2614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8515e-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(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(0.0001, 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(7.2602e-05, device='cuda:0', grad_fn=<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(0.0004, 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(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-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(3.1308e-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(6.3597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, 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(4.0481e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, 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(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-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(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(6.3016e-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(0.0001, 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(0.0004, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-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(0.0008, 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(0.0004, 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(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-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(0.0001, 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(6.9553e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0003, 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(0.0003, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-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(6.0681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0906e-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(0.0005, 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(0.0005, 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(0.0001, 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(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-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(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(0.0007, 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(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(3.3276e-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(8.9556e-05, 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(7.2058e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3954e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6872e-05, device='cuda:0', grad_fn=<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(3.5970e-05, 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(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, 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(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(2.3042e-05, device='cuda:0', grad_fn=<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(7.6248e-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(0.0005, 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(2.5676e-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(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<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(0.0009, 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(0.0005, 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(0.0004, 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(8.2395e-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(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(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-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(3.6908e-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(0.0003, 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(0.0006, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8927e-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(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9941e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-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(0.0006, 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(8.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6813e-05, device='cuda:0', grad_fn=<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(9.8253e-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(0.0001, 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(6.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9951e-05, device='cuda:0', grad_fn=<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(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3765e-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(0.0024, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1998e-05, device='cuda:0', grad_fn=<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(0.0002, 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(1.4382e-05, 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(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8397600685518424, 'agreement': 0.6031746031746031, 'direct_attack': 0.38333333333333336, 'undercutter_attack': 0.32786885245901637, 'partial': 0.14814814814814814}, 'recall': {'support': 0.9486931268151017, 'agreement': 0.2714285714285714, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.07407407407407407}, 'f1': {'support': 0.8909090909090909, 'agreement': 0.37438423645320196, 'direct_attack': 0.2200956937799043, 'undercutter_attack': 0.447427293064877, 'partial': 0.09876543209876543}, 'support': {'support': 1033, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 108}, 'micro_avg': {'precision': 0.6933411214953271, 'recall': 0.6933411214953271, 'f1': 0.6933411214953271, 'support': None}, 'macro_avg': {'precision': 0.4604570011333887, 'recall': 0.4305567081075611, 'f1': 0.4063163492611679, 'support': None}, 'weighted_avg': {'precision': 0.6752541375109749, 'recall': 0.6933411214953271, 'f1': 0.6612922183242246, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-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(1.6562e-05, 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(3.6243e-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(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1415e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9368e-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(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-05, 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(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(0.0006, 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(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-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(9.0404e-05, device='cuda:0', grad_fn=<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(8.8256e-05, device='cuda:0', grad_fn=<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(2.7856e-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(0.0002, 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(0.0003, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(1.4806e-05, device='cuda:0', grad_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(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, 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(0.0001, 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(4.5962e-05, 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(0.0002, 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(5.0472e-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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1772e-05, device='cuda:0', grad_fn=<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(1.5594e-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(1.6441e-05, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-05, device='cuda:0', grad_fn=<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(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-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(0.0006, 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(1.5836e-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(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, 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(0.0004, 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(0.0004, 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(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(0.0008, 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(0.0022, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7074e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8163e-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(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3059e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4985e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<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(3.7121e-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(1.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3200e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(2.1740e-05, device='cuda:0', grad_fn=<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(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-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(0.0006, 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(2.4132e-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(2.9067e-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(0.0003, 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(0.0005, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9932e-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(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(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-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(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4179e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, 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(7.8076e-05, 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(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8949e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-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(0.0001, 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(0.0003, 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(0.0001, 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(8.8826e-05, device='cuda:0', grad_fn=<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(5.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-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(1.2929e-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(6.8424e-05, 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(0.0001, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6853e-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(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(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, 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(6.6456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1371e-05, device='cuda:0', grad_fn=<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(1.4473e-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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-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(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-05, device='cuda:0', grad_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(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(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-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(7.0704e-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(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4800e-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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<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(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-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(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-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(0.0002, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1915e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<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(7.5648e-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(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-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(3.2609e-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(9.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-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(1.5684e-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(0.0003, 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(9.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4993e-05, device='cuda:0', grad_fn=<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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4952e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, 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(1.4049e-05, 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(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(2.1225e-05, 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(0.0005, 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(8.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-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(6.0797e-05, device='cuda:0', grad_fn=<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(1.3020e-05, device='cuda:0', grad_fn=<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(2.5282e-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(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(0.0002, 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(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, 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(1.6926e-05, 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(0.0005, 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(2.8583e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, 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(9.3325e-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(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1563e-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(6.9811e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2798e-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(5.9402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9081e-05, 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(0.0006, 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(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3175e-05, 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(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<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(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(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9139e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8266e-05, device='cuda:0', grad_fn=<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(1.4625e-05, device='cuda:0', grad_fn=<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(0.0019, 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(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4715e-05, 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(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(2.2981e-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(2.7160e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8140e-05, device='cuda:0', grad_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(0.0005, 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(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-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(8.4864e-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(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<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(0.0002, 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>)
			------------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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.839622641509434, 'agreement': 0.6031746031746031, 'direct_attack': 0.3770491803278688, 'undercutter_attack': 0.32786885245901637, 'partial': 0.14814814814814814}, 'recall': {'support': 0.9486434108527132, 'agreement': 0.2714285714285714, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.07339449541284404}, 'f1': {'support': 0.8908098271155597, 'agreement': 0.37438423645320196, 'direct_attack': 0.21904761904761905, 'undercutter_attack': 0.447427293064877, 'partial': 0.098159509202454}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6927570093457944, 'recall': 0.6927570093457944, 'f1': 0.6927570093457944, 'support': None}, 'macro_avg': {'precision': 0.4591726851238141, 'recall': 0.43041084918283745, 'f1': 0.40596569697674234, 'support': None}, 'weighted_avg': {'precision': 0.6742203901827135, 'recall': 0.6927570093457944, 'f1': 0.6606398861837205, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0004, 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(3.5879e-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(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-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(2.7341e-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(1.8500e-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(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8917e-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(5.0662e-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(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0043e-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(1.9499e-05, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2187e-05, 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(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-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(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(1.6017e-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(0.0002, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5497e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9423e-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(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-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(0.0001, device='cuda:0', grad_fn=<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(0.0003, 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(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(9.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1664e-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(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-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(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-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(0.0003, 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(0.0012, 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(5.0411e-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(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5703e-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(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2412e-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(1.6986e-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(0.0003, 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(7.9301e-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(4.1995e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0586e-05, device='cuda:0', grad_fn=<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(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(0.0001, 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(9.2249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<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(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2746e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<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(0.0001, 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(6.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0573e-05, 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(6.1522e-05, 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(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3108e-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(8.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<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(7.2034e-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(2.8219e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-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(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9776e-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(0.0002, 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(6.0201e-05, 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(0.0002, 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(9.7873e-05, device='cuda:0', grad_fn=<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(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, 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(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(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1820e-05, device='cuda:0', grad_fn=<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(9.9725e-05, 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(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7800e-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(0.0001, 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(9.5986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-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(0.0003, 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(0.0004, 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(2.2709e-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(0.0007, 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(0.0009, 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(2.6857e-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(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6718e-05, device='cuda:0', grad_fn=<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(6.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1203e-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(4.1753e-05, 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(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4167e-05, 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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7733e-05, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5446e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-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(0.0005, 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(0.0003, 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(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-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(0.0005, 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(2.8038e-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(4.5597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<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(7.3209e-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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271e-05, device='cuda:0', grad_fn=<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(6.4899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7834e-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(9.5087e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4229e-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(2.0287e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-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(1.3535e-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(9.5912e-05, 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(5.0109e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<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(9.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1964e-05, device='cuda:0', grad_fn=<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(1.4413e-05, 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(9.2669e-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(8.6427e-05, device='cuda:0', grad_fn=<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(8.8261e-05, device='cuda:0', grad_fn=<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(6.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0283e-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(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(6.5076e-05, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5809e-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(5.1260e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-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(2.2618e-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(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(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7273e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-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(1.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-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(3.1580e-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(0.0002, 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(2.0650e-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(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4082e-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(1.6956e-05, 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(2.7220e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7845e-05, 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(0.0002, 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(7.0627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-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(2.2588e-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(0.0006, 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(6.2462e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-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(0.0003, 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(0.0005, 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(5.5834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2621e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-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(0.0002, 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(2.2073e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.839622641509434, 'agreement': 0.6031746031746031, 'direct_attack': 0.3770491803278688, 'undercutter_attack': 0.32786885245901637, 'partial': 0.14814814814814814}, 'recall': {'support': 0.9486434108527132, 'agreement': 0.2714285714285714, 'direct_attack': 0.15436241610738255, 'undercutter_attack': 0.704225352112676, 'partial': 0.07339449541284404}, 'f1': {'support': 0.8908098271155597, 'agreement': 0.37438423645320196, 'direct_attack': 0.21904761904761905, 'undercutter_attack': 0.447427293064877, 'partial': 0.098159509202454}, 'support': {'support': 1032, 'agreement': 280, 'direct_attack': 149, 'undercutter_attack': 142, 'partial': 109}, 'micro_avg': {'precision': 0.6927570093457944, 'recall': 0.6927570093457944, 'f1': 0.6927570093457944, 'support': None}, 'macro_avg': {'precision': 0.4591726851238141, 'recall': 0.43041084918283745, 'f1': 0.40596569697674234, 'support': None}, 'weighted_avg': {'precision': 0.6742203901827135, 'recall': 0.6927570093457944, 'f1': 0.6606398861837205, '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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.5119047619047619, 'agreement': 0.23809523809523808, 'direct_attack': 0.09523809523809523, 'undercutter_attack': 0.15227272727272728, 'partial': 0.0}, 'recall': {'support': 0.10513447432762836, 'agreement': 0.10204081632653061, 'direct_attack': 0.5142857142857142, 'undercutter_attack': 0.34183673469387754, 'partial': 0.0}, 'f1': {'support': 0.1744421906693712, 'agreement': 0.14285714285714285, 'direct_attack': 0.1607142857142857, 'undercutter_attack': 0.21069182389937108, 'partial': 0.0}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.1691988950276243, 'recall': 0.1691988950276243, 'f1': 0.1691988950276243, 'support': None}, 'macro_avg': {'precision': 0.1995021645021645, 'recall': 0.21265954792675013, 'f1': 0.1377410886280342, 'support': None}, 'weighted_avg': {'precision': 0.35123173327593216, 'recall': 0.1691988950276243, 'f1': 0.16194013083689388, 'support': None}}
Loss: tensor(0.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2781, 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
				 {'precision': {'support': 0.725748502994012, 'agreement': 0.5, 'direct_attack': 0.14906832298136646, 'undercutter_attack': 0.16666666666666666, 'partial': 0.18181818181818182}, 'recall': {'support': 0.73992673992674, 'agreement': 0.27692307692307694, 'direct_attack': 0.34285714285714286, 'undercutter_attack': 0.07142857142857142, 'partial': 0.1836734693877551}, 'f1': {'support': 0.732769044740024, 'agreement': 0.3564356435643565, 'direct_attack': 0.2077922077922078, 'undercutter_attack': 0.1, 'partial': 0.182741116751269}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.511049723756906, 'recall': 0.511049723756906, 'f1': 0.511049723756906, 'support': None}, 'macro_avg': {'precision': 0.34466033489204545, 'recall': 0.32296180010465725, 'f1': 0.31594760256957144, 'support': None}, 'weighted_avg': {'precision': 0.527101130976751, 'recall': 0.511049723756906, 'f1': 0.5084546523961759, 'support': None}}
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4460, 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
				 {'precision': {'support': 0.7435294117647059, 'agreement': 0.332541567695962, 'direct_attack': 0.22535211267605634, 'undercutter_attack': 0.3939393939393939, 'partial': 0.0}, 'recall': {'support': 0.7716727716727717, 'agreement': 0.717948717948718, 'direct_attack': 0.22857142857142856, 'undercutter_attack': 0.0663265306122449, 'partial': 0.0}, 'f1': {'support': 0.7573397243858598, 'agreement': 0.4545454545454546, 'direct_attack': 0.22695035460992907, 'undercutter_attack': 0.11353711790393013, 'partial': 0.0}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5642265193370166, 'recall': 0.5642265193370166, 'f1': 0.5642265193370166, 'support': None}, 'macro_avg': {'precision': 0.3390724972152236, 'recall': 0.35690388976103266, 'f1': 0.3104745302890347, 'support': None}, 'weighted_avg': {'precision': 0.5404403390350663, 'recall': 0.5642265193370166, 'f1': 0.5268811620600437, 'support': None}}
Loss: tensor(1.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, 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(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5632, 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(1.8930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, 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(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, 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(0.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, 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(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, 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(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, 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(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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
				 {'precision': {'support': 0.791025641025641, 'agreement': 0.5051546391752577, 'direct_attack': 0.20300751879699247, 'undercutter_attack': 0.26704545454545453, 'partial': 0.125}, 'recall': {'support': 0.7533577533577533, 'agreement': 0.5, 'direct_attack': 0.38571428571428573, 'undercutter_attack': 0.24102564102564103, 'partial': 0.04081632653061224}, 'f1': {'support': 0.7717323327079425, 'agreement': 0.5025641025641026, 'direct_attack': 0.2660098522167487, 'undercutter_attack': 0.2533692722371968, 'partial': 0.061538461538461535}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 195, 'partial': 98}, 'micro_avg': {'precision': 0.5662983425414365, 'recall': 0.5662983425414365, 'f1': 0.5662983425414365, 'support': None}, 'macro_avg': {'precision': 0.37824665070866914, 'recall': 0.38418280132565846, 'f1': 0.37104280425289043, 'support': None}, 'weighted_avg': {'precision': 0.579837862946335, 'recall': 0.5662983425414365, 'f1': 0.5685293516697075, 'support': None}}
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6844, 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(0.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, 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(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, 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(0.1611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, 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(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, 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(1.0732, 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(0.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, 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(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, 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(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3256, 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(1.2952, 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(0.0054, 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(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.3611, 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(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, 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(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6221, 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(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, 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(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5263, 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(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, 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(0.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4357, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, 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(1.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, 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(0.5690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0866, 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(0.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, 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(0.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, 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(0.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3824, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, 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(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0017, 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(0.1215, 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(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5505, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, 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(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, 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(0.0385, 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(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.4810, 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(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, 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(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, 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(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, 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(0.0575, 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(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, 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(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, 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(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, 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(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4685, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, 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(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, 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(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9922, 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(0.6917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4012, 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(0.0150, 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(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, 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(0.3377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5331, 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(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, 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(1.3429, 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(0.0414, 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(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4044, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, 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(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, 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(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, 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(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, 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(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, 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(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, 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(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, 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(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, 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
				 {'precision': {'support': 0.7110423116615067, 'agreement': 0.3938356164383562, 'direct_attack': 0.2676056338028169, 'undercutter_attack': 0.5, 'partial': 0.175}, 'recall': {'support': 0.8422982885085575, 'agreement': 0.5867346938775511, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.09183673469387756, 'partial': 0.14285714285714285}, 'f1': {'support': 0.7711247901510913, 'agreement': 0.4713114754098361, 'direct_attack': 0.18009478672985782, 'undercutter_attack': 0.15517241379310345, 'partial': 0.15730337078651685}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5904696132596685, 'recall': 0.5904696132596685, 'f1': 0.5904696132596685, 'support': None}, 'macro_avg': {'precision': 0.4094967123805359, 'recall': 0.3598882291302829, 'f1': 0.3470013673740811, 'support': None}, 'weighted_avg': {'precision': 0.560386174373912, 'recall': 0.5904696132596685, 'f1': 0.5484806084990522, 'support': None}}
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, 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(0.0183, 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(0.1113, 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(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, 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(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0462, 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(0.0040, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, 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(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0301, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, 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(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1689, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2259, 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(0.0256, 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(0.0052, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, 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(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5118, 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(0.0020, 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(0.1281, 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(0.0013, 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(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, 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(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.1605, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, 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(0.0184, 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(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(0.0438, 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(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0035, 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(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, 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(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7854, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, 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(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, 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(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, 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(0.0086, 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(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(0.0007, 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(0.0208, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4695, 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(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5604, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, 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(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, 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(0.0091, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0617, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, 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(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, 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(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, 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(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7061, 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(0.0021, 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(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7312, 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(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, 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(0.1695, 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(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4194, 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(0.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, 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(0.0043, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1149, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6057, 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(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, 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(0.0016, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7706, 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(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, 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(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0036, 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(0.0029, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, 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(0.0447, 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(0.0015, 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(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, 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(0.1335, 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(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, 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(0.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4190, 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(0.0017, 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(0.0100, 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(0.0156, 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(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7775, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, 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(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4027, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, 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(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, 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(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, 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(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, 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(0.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, 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(0.0327, 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(1.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, 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(0.1910, 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(0.0024, 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(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, 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(0.0152, 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(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, 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(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.3528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, 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(0.0013, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0296, 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(0.0072, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1585, 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(0.0024, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, 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(0.0007, 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(0.0210, 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(0.0059, 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(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, 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(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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
				 {'precision': {'support': 0.6854764107308048, 'agreement': 0.5887096774193549, 'direct_attack': 0.2261904761904762, 'undercutter_attack': 0.35135135135135137, 'partial': 0.24705882352941178}, 'recall': {'support': 0.9058679706601467, 'agreement': 0.37244897959183676, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.1326530612244898, 'partial': 0.21428571428571427}, 'f1': {'support': 0.7804107424960505, 'agreement': 0.45625000000000004, 'direct_attack': 0.16964285714285712, 'undercutter_attack': 0.19259259259259262, 'partial': 0.22950819672131145}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.6077348066298343, 'recall': 0.6077348066298343, 'f1': 0.6077348066298343, 'support': None}, 'macro_avg': {'precision': 0.41975734784427987, 'recall': 0.35219400229529463, 'f1': 0.36568087779056235, 'support': None}, 'weighted_avg': {'precision': 0.5530732714015233, 'recall': 0.6077348066298343, 'f1': 0.5606291013733466, 'support': None}}
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, 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(0.0009, 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(0.0052, 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(0.0007, 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(0.0012, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, 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(0.0151, 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(0.0394, 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(0.0035, 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(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0004, 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(0.0595, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, 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(0.0012, 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(0.0117, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, 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(0.0072, 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(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0123, 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(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0006, 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(0.0009, 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(0.0006, 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(0.0012, 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(0.0013, 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(0.0040, 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(0.0026, 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(0.0019, 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(0.0294, 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(0.0007, 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(0.0035, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, 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(0.0136, 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(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, 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(0.0031, 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(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, 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(0.0081, 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(0.0037, 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(0.0007, 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(0.0035, 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(0.0027, 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(0.0096, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, 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(0.0082, 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(0.0128, 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(0.9375, 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(0.0003, 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(0.0064, 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(0.0141, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0056, 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(0.0140, 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(0.0073, 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(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, 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(0.0507, 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(0.0006, 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(0.1246, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, 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(0.0016, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, 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(0.0006, 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(0.0049, 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(0.0084, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.0107, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0012, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0542, 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(0.0015, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0023, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, 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(0.0024, 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(0.0011, 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(0.0038, 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(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, 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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, 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(0.0786, 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(1.8058, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0051, 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(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0027, 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(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, 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(0.0633, 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(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3537, 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(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, 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(0.0048, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, 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(0.0036, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, 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(0.0015, 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(0.0071, 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(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, 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(0.0060, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, 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(0.0015, 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(0.0602, 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(0.0049, 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(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(0.0065, 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(0.0063, 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(0.0078, 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(0.0139, 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(0.0115, 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(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0029, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, 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(0.0004, 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(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4854, 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(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0055, 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(0.0012, 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(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, 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(0.0027, 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(0.0006, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0049, 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(0.1012, 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(0.0023, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0112, 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(0.0013, 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(0.0030, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7733, 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(0.0081, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0014, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, 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(0.0009, 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(0.0052, 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(0.0013, 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(0.0067, 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(0.0076, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0009, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0123, 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(0.0072, 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(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, 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(0.0056, 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(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(0.0713, 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(0.0054, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0113, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, 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(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, 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(0.0074, 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(0.0007, 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(0.0097, 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(0.0021, 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(0.0019, 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(0.0030, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, 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(0.0325, 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(0.0321, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0002, 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(0.0004, 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(0.0028, 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(0.0053, 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(0.0014, 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(0.0020, 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(0.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.0004, 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(0.0025, 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(0.0053, 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(0.0139, 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(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8882, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, 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(0.0108, 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(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, 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(0.0005, 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(0.0496, 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(0.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, 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(0.0258, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, 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(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, 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(0.6163, device='cuda:0', grad_fn=<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(0.0128, 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(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0063, 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(0.0864, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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
				 {'precision': {'support': 0.6991321118611379, 'agreement': 0.5277777777777778, 'direct_attack': 0.2441860465116279, 'undercutter_attack': 0.30275229357798167, 'partial': 0.2777777777777778}, 'recall': {'support': 0.8852258852258852, 'agreement': 0.3877551020408163, 'direct_attack': 0.1510791366906475, 'undercutter_attack': 0.1683673469387755, 'partial': 0.20408163265306123}, 'f1': {'support': 0.78125, 'agreement': 0.44705882352941173, 'direct_attack': 0.18666666666666665, 'undercutter_attack': 0.21639344262295082, 'partial': 0.23529411764705882}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 139, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.6042817679558011, 'recall': 0.6042817679558011, 'f1': 0.6042817679558011, 'support': None}, 'macro_avg': {'precision': 0.41032520150126056, 'recall': 0.3593018207098371, 'f1': 0.3733326100932176, 'support': None}, 'weighted_avg': {'precision': 0.5500947350050686, 'recall': 0.6042817679558011, 'f1': 0.5655289256643242, 'support': None}}
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, 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(0.0047, 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(0.0013, 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(0.0057, 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(0.0054, 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(0.0005, 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(0.0019, 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(3.9776, 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(0.0081, 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(0.0054, 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(0.0006, 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(0.0118, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, 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(0.0135, 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(0.0025, 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(0.0006, 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(0.0095, 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(0.0076, 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(4.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, 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(0.2389, device='cuda:0', grad_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(0.0048, 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(1.9036, 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(0.3527, 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(0.0117, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, 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(0.0022, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.0057, 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(0.0003, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0002, 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(0.0003, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1391, 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(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, 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(0.0004, 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(0.0094, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, 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(0.0002, 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(0.0020, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, 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(0.0010, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, 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(0.0030, 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(0.0100, 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(0.0114, 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(0.1941, 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(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, 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(0.0391, 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(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, 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(0.0113, 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(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, 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(0.0013, 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(0.0029, 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(0.0089, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, 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(0.0013, 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(0.0053, 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(0.0013, 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(0.0048, 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(0.0004, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, 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(0.0019, 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(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, 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(0.0005, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, 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(0.0074, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, 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(0.0005, 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(0.0015, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, 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(0.0007, 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(0.0050, 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(0.0014, 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(0.0003, 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(0.0056, 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(0.0072, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0008, 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(0.0004, 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(0.0023, 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(0.0017, 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(0.0121, 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(0.0020, 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(0.0009, 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(0.0442, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, 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(0.0063, 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(0.0042, 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(0.0017, 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(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(0.0006, 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(0.0099, 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(0.0071, 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(0.0186, 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(0.0014, 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(0.0044, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0016, 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(0.0010, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0002, 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(0.0013, 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(0.0010, 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(0.0101, 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(0.0223, 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(0.0006, 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(0.0003, 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(0.0042, 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(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, 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(8.3806e-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(0.0031, 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(0.0056, 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(0.0166, 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(0.0026, 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(0.0031, 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(0.0044, 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(0.0008, 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(0.0049, 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(0.0079, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, 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(0.0025, 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(0.0002, 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(0.0016, 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(0.0197, 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(0.0001, 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(0.0112, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, 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(0.0043, 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(0.0004, 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(0.0038, 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(0.0049, 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(0.0006, 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(0.0010, 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(0.0033, 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(0.0019, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, 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(0.0022, 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(0.0006, 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(0.0087, 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(0.0483, device='cuda:0', grad_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(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0053, 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(0.0020, 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(0.0035, 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(0.0006, 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(0.0028, 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(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, 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(0.0020, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, 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(0.0044, 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(0.0025, 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(0.0060, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0003, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, 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(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(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, 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(0.0004, 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(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, 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(0.0049, 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(0.0015, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, 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(0.0006, 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(0.0003, 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(0.0013, 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(0.0055, 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(0.0003, 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(0.0001, 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(0.0014, 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(0.0010, 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(0.0004, 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(0.0105, 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(0.0027, 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(0.0052, 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(0.0054, 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(0.0019, 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(0.0008, 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(0.0025, 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(0.0479, device='cuda:0', grad_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(0.0050, 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(0.0014, 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(0.0056, device='cuda:0', grad_fn=<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(0.0007, 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(0.0002, 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(0.0008, 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(0.0042, 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(0.0024, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, 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(0.0013, 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(0.0033, 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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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(0.0026, 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(0.0022, 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(0.0293, 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(0.0656, 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(0.0004, 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(0.0101, 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(0.0006, 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(0.0011, 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(0.0013, 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(0.0044, 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(0.1494, 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(0.0004, 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(0.0011, 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(8.4986e-05, 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(0.0001, 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(0.0009, 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(0.0038, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, 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(0.0006, 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(0.0334, 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(0.0115, device='cuda:0', grad_fn=<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(0.0031, 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(0.0104, 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(0.0005, 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(0.0012, 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(0.0038, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, 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(0.0019, 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(0.0006, 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(0.0073, 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(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, 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
				 {'precision': {'support': 0.7104477611940299, 'agreement': 0.4476744186046512, 'direct_attack': 0.23893805309734514, 'undercutter_attack': 0.31868131868131866, 'partial': 0.23880597014925373}, 'recall': {'support': 0.8717948717948718, 'agreement': 0.39285714285714285, 'direct_attack': 0.19285714285714287, 'undercutter_attack': 0.14871794871794872, 'partial': 0.16326530612244897}, 'f1': {'support': 0.7828947368421053, 'agreement': 0.4184782608695652, 'direct_attack': 0.2134387351778656, 'undercutter_attack': 0.20279720279720276, 'partial': 0.1939393939393939}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 195, 'partial': 98}, 'micro_avg': {'precision': 0.5959944751381215, 'recall': 0.5959944751381215, 'f1': 0.5959944751381215, 'support': None}, 'macro_avg': {'precision': 0.3909095043453197, 'recall': 0.353898482469911, 'f1': 0.3623096659252265, 'support': None}, 'weighted_avg': {'precision': 0.5446119282565844, 'recall': 0.5959944751381215, 'f1': 0.5605286372103144, 'support': None}}
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(0.0077, 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(0.0013, 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(0.0001, 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(0.0078, device='cuda:0', grad_fn=<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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0037, 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(0.0860, 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(0.0027, 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(0.0085, 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(0.0041, 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(0.0009, 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(0.0003, 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(0.0029, 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(0.0025, 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(0.0004, 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(0.0087, 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(0.0006, 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(0.0032, 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(0.0038, 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(0.0012, 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(0.0012, 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(0.0009, 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(0.0024, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, 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(0.0038, 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(0.0018, 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(0.0005, 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(0.0062, 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(0.0037, 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(0.0027, 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(0.0062, 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(0.0006, 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(0.1443, 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(0.0003, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0011, 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(0.0019, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, 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(0.0072, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0002, 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(0.0022, 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(0.0072, 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(0.0064, 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(0.0037, 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(0.0075, 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(0.0012, 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(0.0002, 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(0.0003, 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(0.0151, 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(0.0077, 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(0.0024, 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(0.0006, 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(0.0001, 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(0.0040, 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(0.0071, 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(0.0030, 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(0.0014, 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(0.0004, 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(0.0379, 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(0.0068, 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(0.0062, 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(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0578, 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(0.0003, 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(0.0237, device='cuda:0', grad_fn=<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(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, 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(0.0005, 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(0.0198, 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(0.0008, 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(0.0053, 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(0.0471, 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(0.0007, 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(0.0026, 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(0.0036, 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(0.0016, 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(0.0373, 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(0.0041, 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(0.0013, 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(0.0026, 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(0.0003, 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(0.0079, 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(0.0003, 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(0.0069, 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(0.0011, 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(0.0117, 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(0.0024, 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(0.0008, 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(0.0017, 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(0.0008, 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(0.0036, 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(0.0002, 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(0.0013, 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(0.0005, 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(0.0017, 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(0.0009, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0194, 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(0.0012, 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(0.0036, 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(0.0003, 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(0.0014, 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(0.0005, 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(8.4472e-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(0.0009, 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(0.0010, 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(0.0011, 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(0.0002, 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(0.0066, 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(0.0040, 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(0.0046, 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(0.0001, 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(0.0036, 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(0.0016, 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(0.0021, 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(0.0002, 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(0.0006, 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(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0005, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0028, 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(0.0021, 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(0.0019, 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(0.0064, 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(0.0016, 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(0.0042, 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(0.0051, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, 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(0.0002, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0009, 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(0.0006, 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(0.0039, 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(0.0013, 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(0.0009, 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(0.0048, 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(0.0004, 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(0.0002, 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(0.0261, 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(0.0068, device='cuda:0', grad_fn=<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(0.0021, 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(0.0033, 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(0.0031, 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(0.0036, 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(0.0041, 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(0.0004, 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(0.0011, 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(0.0071, 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(0.0024, 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(0.0005, 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(0.0004, 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(0.0005, 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(0.0123, device='cuda:0', grad_fn=<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(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4098e-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(0.0001, 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(0.0041, 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(0.0023, 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(0.0023, 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(0.0013, 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(0.0057, 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(0.0001, 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(0.0006, 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(7.0939e-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(0.0017, 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(0.0014, 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(0.0177, 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(0.0005, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, 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(0.0005, 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(0.0017, 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(0.0019, 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(0.0013, 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(0.0022, 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(0.0004, 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(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(0.0039, 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(0.0019, 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(0.0003, 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(0.0034, 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(0.0017, 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(0.0024, 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(0.0026, 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(0.0027, 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(0.0010, 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(0.0133, 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(0.0049, 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(0.0234, 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(0.0061, 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(0.0015, 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(0.0013, 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(0.0003, 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(7.2423e-05, 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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, 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(0.0002, 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(0.0025, 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(0.0016, 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(0.0102, 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(0.0026, 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(0.0006, 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(0.0029, 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(0.0008, 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(0.0113, 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(0.0007, 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(0.0003, 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(0.0015, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<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(8.6257e-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(0.0019, 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(0.0033, 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(0.0021, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2150e-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(0.0002, 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(0.0022, 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(0.0030, 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(0.0033, 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(0.0011, 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(0.0084, 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(0.0003, 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(0.0011, 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(0.0001, 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(0.0023, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5066e-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(0.0003, 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(0.0040, 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(0.0044, 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(0.0004, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6823e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5087e-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(0.0008, 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(0.0004, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, 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(0.0022, 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(0.0002, 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(0.0009, 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(0.0092, device='cuda:0', grad_fn=<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(0.0015, 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(0.0006, 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(0.0014, 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(0.0004, 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(0.0033, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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
				 {'precision': {'support': 0.7068273092369478, 'agreement': 0.4807692307692308, 'direct_attack': 0.21782178217821782, 'undercutter_attack': 0.3333333333333333, 'partial': 0.2345679012345679}, 'recall': {'support': 0.8595848595848596, 'agreement': 0.3826530612244898, 'direct_attack': 0.15827338129496402, 'undercutter_attack': 0.19387755102040816, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7757575757575759, 'agreement': 0.42613636363636365, 'direct_attack': 0.18333333333333332, 'undercutter_attack': 0.24516129032258058, 'partial': 0.2122905027932961}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 139, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5925414364640884, 'recall': 0.5925414364640884, 'f1': 0.5925414364640884, 'support': None}, 'macro_avg': {'precision': 0.3946639113504595, 'recall': 0.3576532808290259, 'f1': 0.36853581316862993, 'support': None}, 'weighted_avg': {'precision': 0.5467683362382063, 'recall': 0.5925414364640884, 'f1': 0.5616074567185664, 'support': None}}
Loss: tensor(0.0005, 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(0.0004, 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(0.0004, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7973e-05, 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(0.0013, 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(0.0010, 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(0.0006, 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(0.0011, 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(0.0034, 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(0.0012, 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(0.0008, 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(0.0015, 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(0.0007, 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(0.0002, 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(0.0001, 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(0.0008, 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(0.0005, 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(0.0005, 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(0.0022, 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(0.0023, 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(0.0019, 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(0.0006, 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(0.0004, 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(0.0001, 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(0.0011, 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(9.0647e-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(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(0.0002, 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(0.0015, 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(0.0011, 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(0.0002, 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(0.0038, 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(0.0004, 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(0.0009, 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(0.0034, 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(0.0026, 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(0.0030, 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(0.0057, 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(0.0019, 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(8.6046e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4542e-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(8.6379e-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(0.0103, 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(0.0014, 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(0.0014, 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(0.0012, 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(0.0008, 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(0.0002, 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(0.0003, 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(6.6791e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3927e-05, 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(0.0016, 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(0.0024, 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(0.0039, 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(0.0002, 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(0.0001, 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(9.7096e-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(0.0101, 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(0.0013, 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(0.0021, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3072, device='cuda:0', grad_fn=<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(0.0017, 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(0.0028, 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(0.0025, 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(0.0010, 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(0.0001, 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(0.0023, 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(0.0003, 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(0.0004, 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(0.0001, 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(0.0032, 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(0.0014, 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(0.0041, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4008e-05, 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(0.0047, 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(0.0005, 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(0.0001, 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(0.0151, 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(8.6530e-05, 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(0.0033, 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(0.0042, device='cuda:0', grad_fn=<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(0.0009, 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(8.9043e-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(0.0012, 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(0.0027, 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(0.0012, 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(9.0313e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, 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(0.0002, 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(0.0016, 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(0.0052, 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(0.0072, 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(0.0005, 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(0.0004, 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(0.0068, 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(0.0003, 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(0.0097, device='cuda:0', grad_fn=<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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, 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(0.0021, 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(0.0019, 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(0.0006, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2039e-05, 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(0.0016, 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(0.0004, 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(0.0041, 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(0.0002, 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(0.0004, 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(0.0004, 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(8.6863e-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(0.0007, 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(0.0009, 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(0.0036, 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(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(0.0076, 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(0.0003, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2826e-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(0.0070, device='cuda:0', grad_fn=<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(0.0002, 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(0.0030, 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(0.0028, 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(0.0029, 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(0.0004, 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(0.0002, 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(0.0033, 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(0.0001, 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(0.0023, 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(0.0013, 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(0.0003, 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(0.0032, 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(0.0042, 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(0.0002, 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(0.0018, 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(0.0004, 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(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(0.0022, 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(0.0009, 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(0.0027, 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(0.0003, 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(0.0011, 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(0.0048, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, 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(0.0003, 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(0.0049, 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(0.0078, 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(0.0042, 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(0.0016, 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(0.0006, 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(0.0182, 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(0.0013, 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(0.0018, 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(8.5652e-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(0.0002, 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(0.0078, 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(0.0012, 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(6.3067e-05, 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(0.0044, 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(0.0003, 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(0.0004, 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(0.0092, device='cuda:0', grad_fn=<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(0.0024, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9788e-05, device='cuda:0', grad_fn=<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(8.4108e-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(0.0002, 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(0.0014, 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(0.0004, 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(0.0055, 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(9.1403e-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(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(0.0116, device='cuda:0', grad_fn=<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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1141e-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(0.0016, 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(0.0011, 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(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(0.0029, 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(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7974e-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(0.0004, 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(0.0002, 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(0.0028, device='cuda:0', grad_fn=<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(0.0082, 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(0.0013, 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(0.0001, 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(0.0004, 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(0.0028, 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(0.0018, 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(0.0064, 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(6.2189e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8820e-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(0.0013, 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(0.0004, 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(0.0002, 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(0.0004, 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(0.0012, 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(0.0003, 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(0.0014, 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(8.8589e-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(0.0004, 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(0.0020, 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(0.0003, 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(0.0027, 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(0.0016, 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(0.0014, 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(0.0016, 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(0.0004, 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(0.0021, 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(0.0048, device='cuda:0', grad_fn=<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(0.0003, 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(0.0002, 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(0.0003, 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(0.0044, 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(6.6761e-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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2987e-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(7.3603e-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(0.0017, 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(0.0003, 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(0.0002, 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(0.0017, 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(6.2583e-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(0.0018, 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(0.0009, 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(0.0031, device='cuda:0', grad_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(0.0010, 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(0.0035, 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(0.0027, 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(0.0004, 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(0.0007, 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(0.0011, 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(0.0012, 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(0.0032, 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(7.2998e-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(0.0032, 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(0.0002, 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(0.0007, 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(0.0009, 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(0.0001, 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(0.0025, 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(0.0004, 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(0.0035, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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
				 {'precision': {'support': 0.7095477386934673, 'agreement': 0.46706586826347307, 'direct_attack': 0.22115384615384615, 'undercutter_attack': 0.36666666666666664, 'partial': 0.2608695652173913}, 'recall': {'support': 0.863080684596577, 'agreement': 0.3979591836734694, 'direct_attack': 0.16428571428571428, 'undercutter_attack': 0.1683673469387755, 'partial': 0.24489795918367346}, 'f1': {'support': 0.7788196359624932, 'agreement': 0.4297520661157025, 'direct_attack': 0.1885245901639344, 'undercutter_attack': 0.23076923076923075, 'partial': 0.25263157894736843}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5966850828729282, 'recall': 0.5966850828729282, 'f1': 0.5966850828729282, 'support': None}, 'macro_avg': {'precision': 0.4050607369989689, 'recall': 0.3677181777356419, 'f1': 0.3760994203917458, 'support': None}, 'weighted_avg': {'precision': 0.5527267838055293, 'recall': 0.5966850828729282, 'f1': 0.5647016393415464, 'support': None}}
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(0.0028, 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(0.0017, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5126e-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(0.0003, 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(0.0019, 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(0.0087, 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(0.0001, 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(0.0008, 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(0.0017, 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(0.0004, 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(0.0040, 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(0.0020, 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(0.0028, 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(0.0022, 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(0.0016, 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(0.0009, 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(0.0013, 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(0.0025, 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(0.0008, 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(0.0008, 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(0.0027, 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(0.0002, 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(0.0006, 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(0.0043, 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(0.0005, 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(0.0006, 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(0.0051, 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(0.0002, 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(0.0005, 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(0.0017, 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(0.0016, 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(0.0005, 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(0.0015, 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(0.0006, 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(6.2008e-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(0.0015, 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(0.0010, 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(0.0001, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9698e-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(0.0014, 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(7.7387e-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(0.0009, 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(0.0002, 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(0.0016, 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(7.9264e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9397e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5137e-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(7.9446e-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(0.0001, 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(0.0008, 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(0.0005, 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(0.0003, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-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(0.0017, 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(0.0010, 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(0.0015, 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(8.3170e-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(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(0.0010, 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(0.0021, 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(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(0.0007, 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(0.0002, 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(0.0021, 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(0.0008, 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(0.0025, 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(0.0022, 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(0.0018, 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(0.0002, 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(0.0002, 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(0.0007, 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(0.0024, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3068e-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(0.0038, 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(0.0018, 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(0.0002, 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(0.0013, 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(0.0003, 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(0.0006, 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(0.0003, 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(0.0012, 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(0.0004, 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(0.0009, 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(0.0007, 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(0.0021, device='cuda:0', grad_fn=<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(0.0020, 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(0.0021, 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(0.0022, 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(0.0022, 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(0.0012, 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(0.0014, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-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(0.0044, device='cuda:0', grad_fn=<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(0.0015, 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(0.0017, 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(0.0003, 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(7.2876e-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(6.4309e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8436e-05, device='cuda:0', grad_fn=<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(0.0012, 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(0.0007, 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(0.0012, 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(0.0023, 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(0.0009, 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(0.0013, 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(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(0.0014, 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(0.0002, 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(0.0002, 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(0.0014, 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(0.0006, 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(0.0023, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-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(0.0001, 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(0.0001, 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(0.0004, 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(0.0014, 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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8225e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-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(8.3352e-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(0.0007, 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(0.0002, 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(0.0001, 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(0.0001, 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(0.0012, 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(0.0019, 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(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1827e-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(0.0025, 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(0.0003, 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(0.0001, 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(0.0002, 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(0.0006, 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(0.0023, 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(0.0032, 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(0.0026, 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(0.0001, 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(0.0030, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7487e-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(0.0004, 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(0.0010, 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(7.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-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(8.8224e-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(0.0063, device='cuda:0', grad_fn=<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(0.0007, 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(0.0016, 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(0.0002, 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(0.0009, 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(8.7679e-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(0.0011, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-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(4.6204e-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(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(0.0027, 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(0.0002, 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(0.0006, 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(0.0019, 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(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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4179e-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(0.0019, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4220e-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(0.0013, 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(0.0005, 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(0.0004, 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(0.0017, 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(6.3037e-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(0.0022, 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(5.2774e-05, device='cuda:0', grad_fn=<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(0.0042, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4432e-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(0.0009, 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(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(0.0002, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5885e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0022, 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(0.0048, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4502e-05, device='cuda:0', grad_fn=<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(0.0022, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7206e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6367e-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(0.0008, 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(0.0015, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8800e-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(0.0009, 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(9.3341e-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(0.0014, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2009e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6871e-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(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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-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(0.0014, 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(6.2402e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2474e-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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0989e-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(0.0012, 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(0.0017, 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(0.0004, 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(0.0021, 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(0.0015, 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(4.8777e-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(9.5158e-05, 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(0.0025, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-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(0.0005, 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(0.0078, 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(0.0006, 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(0.0019, 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(0.0014, 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(0.0005, 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(0.0008, 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>)
			------------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
				 {'precision': {'support': 0.7116935483870968, 'agreement': 0.46706586826347307, 'direct_attack': 0.2079207920792079, 'undercutter_attack': 0.34579439252336447, 'partial': 0.24691358024691357}, 'recall': {'support': 0.8620268620268621, 'agreement': 0.4, 'direct_attack': 0.15, 'undercutter_attack': 0.18877551020408162, 'partial': 0.20408163265306123}, 'f1': {'support': 0.7796797349530646, 'agreement': 0.430939226519337, 'direct_attack': 0.17427385892116182, 'undercutter_attack': 0.2442244224422442, 'partial': 0.223463687150838}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5953038674033149, 'recall': 0.5953038674033149, 'f1': 0.5953038674033149, 'support': None}, 'macro_avg': {'precision': 0.39587763630001116, 'recall': 0.36097680097680096, 'f1': 0.37051618599732916, 'support': None}, 'weighted_avg': {'precision': 0.549058703819251, 'recall': 0.5953038674033149, 'f1': 0.564058439562331, 'support': None}}
Loss: tensor(0.0018, 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(9.2996e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9467e-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(0.0015, 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(0.0068, 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(0.0005, 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(0.0008, 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(0.0009, 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(0.0004, 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(0.0004, 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(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(0.0011, 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(0.0019, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7376e-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(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0253e-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(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(0.0005, 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(7.8568e-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(0.0015, 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(0.0039, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7984e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0006, 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(6.9606e-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(0.0002, 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(0.0037, 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(0.0002, 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(0.0008, 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(8.1472e-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(0.0022, 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(0.0042, 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(0.0006, 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(0.0002, 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(0.0005, 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(9.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-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(0.0034, device='cuda:0', grad_fn=<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(0.0011, 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(4.9867e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5096e-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(0.0002, 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(0.0025, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3569e-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(0.0002, 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(0.0002, 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(0.0031, 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(0.0016, 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(0.0006, 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(5.4106e-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(0.0002, 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(0.0013, 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(0.0010, 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(0.0011, 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(0.0005, 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(0.0038, device='cuda:0', grad_fn=<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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1826e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-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(0.0001, 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(0.0020, 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(0.0003, 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(0.0021, 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(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3391e-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(9.7883e-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(0.0001, 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(7.2301e-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(0.0001, 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(0.0023, 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(0.0006, 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(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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4844e-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(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(0.0016, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0968e-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(6.3309e-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(0.0002, 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(0.0016, 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(0.0020, 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(0.0024, 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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0001, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2735e-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(0.0002, 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(0.0040, 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(0.0007, 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(0.0012, 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(9.5731e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4764e-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(0.0001, 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(9.2100e-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(5.4287e-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(5.1593e-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(0.0006, 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(0.0016, 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(0.0022, 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(0.0020, 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(0.0002, 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(0.0002, 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(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6075e-05, device='cuda:0', grad_fn=<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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-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(8.3320e-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(0.0026, 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(0.0034, 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(5.3622e-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(0.0012, 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(0.0002, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-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(0.0012, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1343e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3099e-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(0.0021, 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(0.0001, 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(0.0006, 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(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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7063e-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(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(0.0012, 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(0.0002, 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(7.7659e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, 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(6.6972e-05, 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(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(0.0003, 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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3099e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8435e-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(0.0002, 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(0.0013, 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(0.0001, 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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7648e-05, device='cuda:0', grad_fn=<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(8.1898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-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(0.0009, 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(0.0008, 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(8.9467e-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(9.2885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0009, 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(0.0010, 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(0.0015, 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(9.9818e-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(0.0001, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5824e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5340e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0032, 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(0.0003, 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(0.0002, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1263e-05, 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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(8.3925e-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(0.0004, 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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4673e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5439e-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(0.0007, 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(0.0008, 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(0.0002, 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(0.0012, 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(8.6317e-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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-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(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-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(0.0014, 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(0.0027, 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(0.0007, 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(4.7294e-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(0.0012, 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(0.0010, 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(0.0011, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5895e-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(8.3654e-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(0.0013, 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(0.0011, 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(0.0012, 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(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(0.0001, 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(0.0028, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, 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(7.1453e-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(5.1260e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, 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(0.0001, 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(0.0015, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0173e-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(0.0002, 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(0.0011, 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(0.0003, 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(6.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0981e-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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-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(0.0006, 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(0.0009, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4774e-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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-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(0.0056, device='cuda:0', grad_fn=<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(0.0013, 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(0.0016, 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(0.0007, 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(0.0003, 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>)
			------------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
				 {'precision': {'support': 0.7106854838709677, 'agreement': 0.47023809523809523, 'direct_attack': 0.21212121212121213, 'undercutter_attack': 0.3482142857142857, 'partial': 0.2597402597402597}, 'recall': {'support': 0.8618581907090465, 'agreement': 0.4030612244897959, 'direct_attack': 0.15, 'undercutter_attack': 0.1989795918367347, 'partial': 0.20408163265306123}, 'f1': {'support': 0.7790055248618786, 'agreement': 0.43406593406593413, 'direct_attack': 0.17573221757322174, 'undercutter_attack': 0.2532467532467533, 'partial': 0.22857142857142856}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5966850828729282, 'recall': 0.5966850828729282, 'f1': 0.5966850828729282, 'support': None}, 'macro_avg': {'precision': 0.4001998673369641, 'recall': 0.36359612793772766, 'f1': 0.37412437166384327, 'support': None}, 'weighted_avg': {'precision': 0.5503514555418739, 'recall': 0.5966850828729282, 'f1': 0.5655678981426482, 'support': None}}
Loss: tensor(0.0005, 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(0.0015, 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(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-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(8.8013e-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(0.0013, 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(0.0008, 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(0.0008, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7627e-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(0.0009, 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(0.0004, 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(0.0012, 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(0.0006, 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(0.0007, 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(0.0005, 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(8.6166e-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(0.0006, 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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-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(0.0011, 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(0.0006, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8465e-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(6.4460e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4563e-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(0.0009, 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(0.0008, 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(0.0009, 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(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8942e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3481e-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(3.3306e-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(5.7467e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8211e-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(8.5258e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7610e-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(8.6455e-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(0.0002, 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(0.0009, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-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(7.7145e-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(0.0004, 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(8.5167e-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(6.4793e-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(0.0026, 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(0.0020, 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(0.0009, 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(0.0007, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6416e-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(0.0007, 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(0.0001, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9558e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-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(0.0003, 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(0.0001, 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(0.0008, 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(0.0006, 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(0.0008, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6095e-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(9.7338e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8374e-05, 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(0.0011, 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(0.0007, 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(0.0001, 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(0.0002, 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(0.0002, 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(0.0012, 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(0.0010, 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(9.4553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(7.0908e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, 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(0.0021, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, 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(0.0015, 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(0.0025, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0022, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8840e-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(0.0006, 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(0.0013, 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(5.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3825e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0024, device='cuda:0', grad_fn=<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(0.0005, 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(0.0008, 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(0.0002, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<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(9.0737e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(9.1949e-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(9.2827e-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(0.0004, 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(0.0002, 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(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0212e-05, device='cuda:0', grad_fn=<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(6.8334e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3239e-05, device='cuda:0', grad_fn=<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(0.0011, 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(5.9949e-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(8.0322e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6024e-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(6.4642e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8216e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3086e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3390e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0003, 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(0.0021, 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(0.0057, 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(0.0025, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3432e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9717e-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(0.0009, 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(0.0003, 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(9.1433e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9770e-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(0.0001, 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(9.1918e-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(0.0010, 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(0.0007, 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(5.9616e-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(0.0006, 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(0.0032, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2786e-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(0.0006, 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(0.0013, 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(0.0007, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9225e-05, device='cuda:0', grad_fn=<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(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4965e-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(0.0009, 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(7.2241e-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(0.0004, 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(6.0251e-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(7.3634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8507e-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(0.0002, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-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(0.0017, 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(0.0001, 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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-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(4.8565e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5561e-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(0.0006, 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(6.9698e-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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2261e-05, device='cuda:0', grad_fn=<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(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-05, 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(0.0006, 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(0.0008, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7506e-05, 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(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<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(0.0001, 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(7.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6153e-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(7.3663e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(6.6306e-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(7.8386e-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(7.7447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-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(4.5144e-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(0.0039, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6479e-05, 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(0.0030, 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(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-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(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-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(0.0002, 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(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(4.5598e-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(0.0009, 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(8.8800e-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(0.0003, 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(4.3872e-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(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-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(8.4259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0011e-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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8356e-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(0.0001, 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(0.0002, 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(0.0002, 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(0.0003, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5358e-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(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, 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(0.0001, 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>)
			------------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
				 {'precision': {'support': 0.7129909365558912, 'agreement': 0.47023809523809523, 'direct_attack': 0.21212121212121213, 'undercutter_attack': 0.34513274336283184, 'partial': 0.26666666666666666}, 'recall': {'support': 0.8644688644688645, 'agreement': 0.40512820512820513, 'direct_attack': 0.15, 'undercutter_attack': 0.1989795918367347, 'partial': 0.20408163265306123}, 'f1': {'support': 0.7814569536423841, 'agreement': 0.4352617079889807, 'direct_attack': 0.17573221757322174, 'undercutter_attack': 0.25242718446601947, 'partial': 0.23121387283236994}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5987569060773481, 'recall': 0.5987569060773481, 'f1': 0.5987569060773481, 'support': None}, 'macro_avg': {'precision': 0.4014299307889394, 'recall': 0.3645316588173731, 'f1': 0.37521838730059515, 'support': None}, 'weighted_avg': {'precision': 0.5518731535498077, 'recall': 0.5987569060773481, 'f1': 0.5674215996161097, 'support': None}}
Loss: tensor(0.0002, 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(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(5.3228e-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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7741e-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(0.0009, 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(0.0019, 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(0.0010, 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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-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(0.0001, 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(6.3158e-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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0002, 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(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(0.0008, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3461e-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(7.7689e-05, device='cuda:0', grad_fn=<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(9.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8265e-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(0.0002, 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(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(0.0002, device='cuda:0', grad_fn=<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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2877e-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(5.0715e-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(5.9161e-05, 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(0.0009, 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(0.0018, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-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(8.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-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(0.0028, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8226e-05, device='cuda:0', grad_fn=<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(7.6540e-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(0.0003, 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(0.0004, 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(6.9304e-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(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(0.0013, 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(0.0005, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7669e-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(0.0018, 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(0.0008, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6891e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, 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(0.0002, 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(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(0.0021, 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(0.0012, 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(9.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2200e-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(0.0012, 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(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(8.2685e-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(9.3311e-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(9.5669e-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(0.0015, 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(0.0006, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-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(0.0016, 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(8.1497e-05, device='cuda:0', grad_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(3.8816e-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(5.7738e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0008, 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(0.0005, 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(0.0008, 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(0.0013, 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(3.7424e-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(8.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-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(6.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-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(5.4136e-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(0.0017, 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(0.0005, 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(6.5701e-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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0001, 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(6.7124e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, 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(0.0004, 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(0.0001, 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(0.0010, 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(0.0013, 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(0.0004, 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(0.0002, 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(0.0006, 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(0.0005, 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(5.2380e-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(4.7051e-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(6.3369e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7892e-05, 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(7.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-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(5.0654e-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(3.9331e-05, 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(0.0001, 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(9.9426e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5298e-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(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(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-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(0.0002, 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(4.2722e-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(9.4099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0008, 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(0.0011, 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(9.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-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(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-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(8.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0008, 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(6.9334e-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(7.3148e-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(0.0001, 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(6.6367e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, 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(7.7993e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9719e-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(0.0004, 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(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6042e-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(9.5611e-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(8.0021e-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(7.7416e-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(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0839e-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(9.7018e-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(0.0002, 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(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7538e-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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7751e-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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1433e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-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(0.0030, 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(0.0010, 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(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0030e-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(4.9564e-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(0.0011, 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(0.0012, 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(9.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8223e-05, device='cuda:0', grad_fn=<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(4.9140e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8545e-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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9798e-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(0.0011, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9960e-05, device='cuda:0', grad_fn=<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(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<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(3.0157e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6399e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6580e-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(0.0004, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-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(0.0002, 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(0.0005, 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(0.0014, 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(0.0002, device='cuda:0', grad_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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1776e-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(2.9975e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0010, 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(0.0006, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(8.8768e-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(0.0010, 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(9.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9868e-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(7.9899e-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(0.0001, 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(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3592e-05, device='cuda:0', grad_fn=<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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<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(9.7035e-05, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9889e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2262e-05, device='cuda:0', grad_fn=<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(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0001, 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(6.0888e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5489e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6698e-05, device='cuda:0', grad_fn=<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(9.3856e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1917e-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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0005, 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(0.0008, 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(0.0015, 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(6.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9001e-05, 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(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-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(0.0012, 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(5.0442e-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(0.0013, 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(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7194e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.710261569416499, 'agreement': 0.46706586826347307, 'direct_attack': 0.21428571428571427, 'undercutter_attack': 0.34210526315789475, 'partial': 0.26666666666666666}, 'recall': {'support': 0.863080684596577, 'agreement': 0.3979591836734694, 'direct_attack': 0.15, 'undercutter_attack': 0.1989795918367347, 'partial': 0.20408163265306123}, 'f1': {'support': 0.7792494481236203, 'agreement': 0.4297520661157025, 'direct_attack': 0.1764705882352941, 'undercutter_attack': 0.25161290322580643, 'partial': 0.23121387283236994}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5966850828729282, 'recall': 0.5966850828729282, 'f1': 0.5966850828729282, 'support': None}, 'macro_avg': {'precision': 0.4000770163580495, 'recall': 0.36282021855196844, 'f1': 0.37365977570655867, 'support': None}, 'weighted_avg': {'precision': 0.5495337285045702, 'recall': 0.5966850828729282, 'f1': 0.5651508456122726, 'support': None}}
Loss: tensor(2.6342e-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(0.0016, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-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(0.0001, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-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(0.0013, 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(9.6789e-05, device='cuda:0', grad_fn=<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(4.5295e-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(0.0006, 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(0.0008, 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(0.0003, 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(0.0010, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2150e-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(0.0015, 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(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, 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(0.0009, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6650e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8586e-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(7.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9077e-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(9.1734e-05, 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(8.8467e-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(7.5754e-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(0.0006, 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(0.0008, 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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3089e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<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(6.8244e-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(0.0001, 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(0.0005, 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(8.1654e-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(8.6893e-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(0.0001, 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(0.0007, 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(0.0011, 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(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(0.0009, 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(0.0020, 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(6.2159e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-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(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3560e-05, 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(2.7069e-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(4.2995e-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(7.1270e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6027e-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(9.5552e-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(8.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-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(5.2925e-05, 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(6.5035e-05, 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(0.0002, 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(3.1217e-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(0.0003, 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(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-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(2.8734e-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(7.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6508e-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(0.0002, 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(7.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-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(0.0002, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5299e-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(5.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1594e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4489e-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(0.0005, 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(0.0002, 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(0.0005, 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(0.0001, 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(0.0001, 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(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(0.0002, 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(0.0007, 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(0.0010, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3748e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7759e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9705e-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(0.0004, 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(4.6779e-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(0.0012, 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(0.0018, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2764e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-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(9.5790e-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(8.5726e-05, device='cuda:0', grad_fn=<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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, 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(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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8779e-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(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(3.6485e-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(0.0009, 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(0.0001, 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(4.1784e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2877e-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(8.5189e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, 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(7.0546e-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(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-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(0.0003, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9715e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3411e-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(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-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(7.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8366e-05, device='cuda:0', grad_fn=<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(3.3064e-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(5.8163e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1936e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(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(0.0006, 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(8.4682e-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(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-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(9.5912e-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>)
Loss: tensor(3.1641e-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(0.0010, 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(0.0003, 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(5.8829e-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(3.4790e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-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(8.4078e-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(0.0015, 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(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, 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(0.0009, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3169e-05, 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(4.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-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(9.1926e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5429e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-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(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3752e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0045, device='cuda:0', grad_fn=<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(7.9718e-05, device='cuda:0', grad_fn=<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(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4368e-05, 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(0.0001, 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(8.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8599e-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(7.3723e-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(9.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-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(8.7589e-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(9.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-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(6.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5347e-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(0.0005, 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(7.7174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-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(2.7372e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2763e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-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(5.5710e-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(0.0005, 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(8.2534e-05, device='cuda:0', grad_fn=<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(0.0004, 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(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(9.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5610e-05, 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(9.1737e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-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(0.0002, 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(8.5681e-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(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-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(7.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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
				 {'precision': {'support': 0.7112676056338029, 'agreement': 0.4642857142857143, 'direct_attack': 0.21428571428571427, 'undercutter_attack': 0.34210526315789475, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8632478632478633, 'agreement': 0.3979591836734694, 'direct_attack': 0.1510791366906475, 'undercutter_attack': 0.1989795918367347, 'partial': 0.20408163265306123}, 'f1': {'support': 0.7799227799227798, 'agreement': 0.4285714285714286, 'direct_attack': 0.17721518987341775, 'undercutter_attack': 0.25161290322580643, 'partial': 0.23255813953488372}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 139, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5973756906077348, 'recall': 0.5973756906077348, 'f1': 0.5973756906077348, 'support': None}, 'macro_avg': {'precision': 0.40044291352667927, 'recall': 0.36306948162035524, 'f1': 0.37397608822566325, 'support': None}, 'weighted_avg': {'precision': 0.5503128462467077, 'recall': 0.5973756906077348, 'f1': 0.5659506179943635, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-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(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4946e-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(0.0004, 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(4.8444e-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(0.0002, 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(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2552e-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(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2477e-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(5.1078e-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(7.2058e-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(6.0100e-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(7.0364e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<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(6.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-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(9.0463e-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(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-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(0.0005, 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(5.3984e-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(0.0006, 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(0.0003, 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(4.4720e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, 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(9.8488e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7889e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-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(0.0008, 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(7.9353e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6012e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-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(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-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(4.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9720e-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(0.0003, 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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.5619e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-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(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8175e-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(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(0.0006, 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(0.0008, 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(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5844e-05, 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(0.0003, 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(5.1867e-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(6.2129e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-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(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-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(0.0008, 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(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5884e-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(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(0.0004, 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(0.0005, 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(0.0008, 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(0.0004, 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(0.0008, 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(0.0008, 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(0.0005, 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(0.0010, 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(6.2764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4330e-05, device='cuda:0', grad_fn=<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(5.6467e-05, device='cuda:0', grad_fn=<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(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(5.9646e-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(0.0007, 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(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2336e-05, device='cuda:0', grad_fn=<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(0.0005, 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(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9891e-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(4.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4754e-05, 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(0.0006, 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(7.3361e-05, device='cuda:0', grad_fn=<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(4.8323e-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(6.5459e-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(9.7727e-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(7.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5278e-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(5.5589e-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(0.0019, 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(0.0006, 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(0.0003, 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(2.0862e-05, device='cuda:0', grad_fn=<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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7628e-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(0.0008, 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(0.0001, 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(0.0019, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1673e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8095e-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(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8947e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7111e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-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(8.0294e-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(7.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2120e-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(0.0004, 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(8.5225e-05, device='cuda:0', grad_fn=<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(0.0004, 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(3.5607e-05, device='cuda:0', grad_fn=<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(8.3983e-05, 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(0.0013, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-05, 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(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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8760e-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(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(0.0011, 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(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-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(8.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-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(0.0004, 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(5.9645e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-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(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(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1887e-05, device='cuda:0', grad_fn=<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(5.5891e-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(0.0003, 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(6.0857e-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(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-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(6.4944e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4569e-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(4.3025e-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(5.9465e-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(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-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(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-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(0.0001, 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(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-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(0.0006, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-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(5.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9777e-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(7.0182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-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(7.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, 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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(3.4033e-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(0.0006, 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(9.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-05, device='cuda:0', grad_fn=<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(6.2855e-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(0.0005, 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(0.0003, 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(6.1008e-05, device='cuda:0', grad_fn=<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(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0885e-05, device='cuda:0', grad_fn=<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(7.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-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(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1181e-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(0.0003, 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(8.0926e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-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(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8598e-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(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0001, 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(2.1710e-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(6.2279e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9810e-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(4.2722e-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(0.0004, 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(0.0004, 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(0.0001, 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(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9240e-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(0.0010, 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(4.8596e-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(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6901e-05, device='cuda:0', grad_fn=<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(1.9409e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2845e-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(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-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(0.0003, 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(0.0006, 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(0.0004, 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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8417e-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(4.9807e-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(7.5720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4289e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2360e-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(9.0828e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(0.0002, 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(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0010, 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(3.6031e-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(7.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-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>)
			------------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
				 {'precision': {'support': 0.7098393574297188, 'agreement': 0.46107784431137727, 'direct_attack': 0.21428571428571427, 'undercutter_attack': 0.34210526315789475, 'partial': 0.273972602739726}, 'recall': {'support': 0.8632478632478633, 'agreement': 0.39487179487179486, 'direct_attack': 0.15, 'undercutter_attack': 0.1989795918367347, 'partial': 0.20408163265306123}, 'f1': {'support': 0.7790633608815426, 'agreement': 0.425414364640884, 'direct_attack': 0.1764705882352941, 'undercutter_attack': 0.25161290322580643, 'partial': 0.23391812865497078}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5966850828729282, 'recall': 0.5966850828729282, 'f1': 0.5966850828729282, 'support': None}, 'macro_avg': {'precision': 0.4002561563848862, 'recall': 0.3622361765218908, 'f1': 0.3732958691276996, 'support': None}, 'weighted_avg': {'precision': 0.5491509392424716, 'recall': 0.5966850828729282, 'f1': 0.564885829875927, 'support': None}}
Loss: tensor(0.0003, 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(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-05, 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(0.0005, 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(3.9573e-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(9.9679e-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(7.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(2.7553e-05, device='cuda:0', grad_fn=<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(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(8.7012e-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(5.3863e-05, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-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(0.0007, 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(0.0004, 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(7.0090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-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(0.0008, 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(9.2310e-05, device='cuda:0', grad_fn=<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(0.0008, 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(7.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8636e-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(5.2682e-05, 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(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<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(6.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0452e-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(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-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(5.0593e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-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(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(0.0005, 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(2.6524e-05, 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(4.9352e-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(0.0002, device='cuda:0', grad_fn=<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(0.0008, 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(0.0013, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0596e-05, device='cuda:0', grad_fn=<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(7.4783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8938e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-05, device='cuda:0', grad_fn=<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(0.0003, 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(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(8.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-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(5.7527e-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(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2250e-05, 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(0.0004, 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(8.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4044e-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(2.4071e-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(0.0004, 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(9.9094e-05, 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(4.6779e-05, 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(0.0005, 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(5.5225e-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(0.0015, 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(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(0.0007, 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(8.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(9.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-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(0.0015, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(4.6204e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-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(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8642e-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(7.5328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2796e-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(2.7069e-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(0.0002, 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(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-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(8.3076e-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(0.0008, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7024e-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(6.9845e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-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(6.6880e-05, device='cuda:0', grad_fn=<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(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-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(0.0004, 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(4.0784e-05, device='cuda:0', grad_fn=<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(8.7620e-05, device='cuda:0', grad_fn=<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(9.4976e-05, 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(0.0001, 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(8.4384e-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(0.0004, 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(8.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9303e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5344e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6519e-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(0.0002, 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(7.5904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(3.8302e-05, device='cuda:0', grad_fn=<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(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9182e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8151e-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(6.6964e-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(4.9716e-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(0.0002, 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(0.0002, 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(4.2510e-05, 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(2.4162e-05, 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(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1248e-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(8.9739e-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(8.6924e-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(0.0005, 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(5.5983e-05, device='cuda:0', grad_fn=<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(9.6383e-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(4.8989e-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(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(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2257e-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(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-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(6.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-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(9.2648e-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(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(5.2017e-05, device='cuda:0', grad_fn=<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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<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(0.0004, 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(6.8093e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(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(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-05, device='cuda:0', grad_fn=<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(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0006, 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(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9988e-05, device='cuda:0', grad_fn=<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(4.5567e-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(0.0002, 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(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<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(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6843e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-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(8.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-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(0.0011, 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(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8294e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-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(0.0001, 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(5.1956e-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(0.0006, 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(4.4992e-05, 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(8.0381e-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(6.2582e-05, 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(0.0033, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4610e-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(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8082e-05, device='cuda:0', grad_fn=<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(3.5395e-05, device='cuda:0', grad_fn=<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(4.9261e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6711e-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(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7388e-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(5.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, 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(5.9797e-05, device='cuda:0', grad_fn=<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(0.0003, 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(5.8132e-05, device='cuda:0', grad_fn=<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(6.8214e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8345e-05, 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(9.6852e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5941e-05, device='cuda:0', grad_fn=<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(2.0892e-05, device='cuda:0', grad_fn=<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(4.9958e-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(5.4655e-05, device='cuda:0', grad_fn=<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(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0303e-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(2.6039e-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(8.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, 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(0.0011, 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(0.0001, 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(0.0003, 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(0.0007, 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(0.0003, 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(5.4801e-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(6.6336e-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(7.2544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-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(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9724e-05, device='cuda:0', grad_fn=<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(8.7399e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<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(5.0685e-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(6.5903e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7108433734939759, 'agreement': 0.4647058823529412, 'direct_attack': 0.20212765957446807, 'undercutter_attack': 0.3508771929824561, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8644688644688645, 'agreement': 0.4030612244897959, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.20408163265306123, 'partial': 0.20618556701030927}, 'f1': {'support': 0.7801652892561983, 'agreement': 0.43169398907103834, 'direct_attack': 0.16239316239316237, 'undercutter_attack': 0.25806451612903225, 'partial': 0.23391812865497075}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 97}, 'micro_avg': {'precision': 0.5980662983425414, 'recall': 0.5980662983425414, 'f1': 0.5980662983425414, 'support': None}, 'macro_avg': {'precision': 0.39976487573482233, 'recall': 0.36270231486726334, 'f1': 0.37324701710088043, 'support': None}, 'weighted_avg': {'precision': 0.550102965617366, 'recall': 0.5980662983425414, 'f1': 0.5660035498167232, 'support': None}}
Loss: tensor(0.0010, 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(0.0002, 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(3.8393e-05, 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(4.7778e-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(8.3473e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-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(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-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(6.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2462e-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(0.0028, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-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(4.8020e-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(2.8704e-05, 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(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4704e-05, 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(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9010e-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(1.8076e-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(3.9664e-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(7.8658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3068e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-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(4.8656e-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(7.4117e-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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-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(0.0001, 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(0.0007, 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(0.0002, 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(0.0005, 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(0.0003, 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(7.2450e-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(4.2026e-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(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(3.6031e-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(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6880e-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(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6521e-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(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, 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(7.2724e-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(0.0016, 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(0.0006, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8655e-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(0.0007, 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(0.0009, 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(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-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(5.7101e-05, device='cuda:0', grad_fn=<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(9.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(4.4569e-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(0.0003, 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(0.0006, 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(6.8577e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5315e-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(2.6887e-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(7.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0312e-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(7.3239e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, 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(0.0004, 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(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4550e-05, 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(0.0002, 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(5.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-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(0.0001, 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(0.0006, 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(3.3366e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-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(6.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9386e-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(0.0015, 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(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-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(2.5040e-05, 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(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3519e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9248e-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(0.0001, 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(5.2066e-05, 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(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-05, device='cuda:0', grad_fn=<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(0.0016, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0001, device='cuda:0', grad_fn=<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(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-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(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-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(6.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8034e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-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(4.5841e-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(7.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, 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(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6795e-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(5.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4828e-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(9.6440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4245e-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(0.0004, 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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5850e-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(8.0654e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4163e-05, device='cuda:0', grad_fn=<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(0.0009, 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(7.1949e-05, 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(5.6436e-05, 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(6.7848e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-05, device='cuda:0', grad_fn=<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(5.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2534e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2552e-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(5.7254e-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(6.1098e-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(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(0.0001, 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(0.0001, 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(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-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(2.9733e-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(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(5.2501e-05, device='cuda:0', grad_fn=<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(7.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2865e-05, 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(7.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-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(5.2986e-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(0.0005, 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(3.1217e-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(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7226e-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(0.0004, 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(0.0008, 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(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(6.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5265e-05, 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(0.0006, 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(2.2406e-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(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-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(0.0009, 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(5.3686e-05, device='cuda:0', grad_fn=<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(0.0004, 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(8.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(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(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1998e-05, device='cuda:0', grad_fn=<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(3.9845e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-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(6.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6381e-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(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-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(7.1424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7346e-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(9.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, 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(0.0005, 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(7.0030e-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(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(2.6857e-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(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4643e-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(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(2.8219e-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(0.0004, 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(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1041e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2148e-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(0.0008, 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(4.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-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(1.6714e-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(1.6017e-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(4.0814e-05, 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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-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(8.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-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(8.0081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-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(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8557e-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(0.0004, 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(0.0003, 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(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-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(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2411e-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(3.0006e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9172e-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(5.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0681e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7098393574297188, 'agreement': 0.4619883040935672, 'direct_attack': 0.2, 'undercutter_attack': 0.35398230088495575, 'partial': 0.2602739726027397}, 'recall': {'support': 0.8643031784841075, 'agreement': 0.4030612244897959, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.20408163265306123, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7794928335170892, 'agreement': 0.4305177111716621, 'direct_attack': 0.16170212765957445, 'undercutter_attack': 0.2588996763754045, 'partial': 0.2222222222222222}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5966850828729282, 'recall': 0.5966850828729282, 'f1': 0.5966850828729282, 'support': None}, 'macro_avg': {'precision': 0.3972167870021963, 'recall': 0.3602075744723317, 'f1': 0.37056691418919047, 'support': None}, 'weighted_avg': {'precision': 0.5484017142737354, 'recall': 0.5966850828729282, 'f1': 0.5643418656259132, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-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(4.7898e-05, 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(6.5687e-05, 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(0.0006, 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(3.1974e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142e-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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0031e-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(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, 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(2.2224e-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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1430e-05, device='cuda:0', grad_fn=<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(6.3845e-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(0.0007, 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(0.0005, 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(8.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-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(3.6061e-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(0.0006, 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(0.0003, 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(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6860e-05, device='cuda:0', grad_fn=<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(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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5679e-05, 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(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(8.8853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, 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(0.0005, 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(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5480e-05, device='cuda:0', grad_fn=<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(9.7667e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4300e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-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(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4812e-05, 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(8.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0511e-05, device='cuda:0', grad_fn=<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(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(4.1995e-05, 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(3.8725e-05, device='cuda:0', grad_fn=<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(6.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2862e-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(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0010e-05, device='cuda:0', grad_fn=<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(4.6688e-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(9.7187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1339e-05, device='cuda:0', grad_fn=<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(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3166e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-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(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4056e-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(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(2.3829e-05, device='cuda:0', grad_fn=<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(3.5940e-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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5512e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5091e-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(7.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8389e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7004e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-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(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-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(3.1943e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-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(2.8613e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5581e-05, device='cuda:0', grad_fn=<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(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(5.5953e-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(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(3.6243e-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(8.6725e-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(0.0005, 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(7.2855e-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(3.4820e-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(7.2754e-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(0.0005, 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(9.1182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8059e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-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(2.9219e-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(0.0005, 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(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1968e-05, 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(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8180e-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(3.7605e-05, device='cuda:0', grad_fn=<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(5.2773e-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(0.0001, 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(0.0001, 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(3.9301e-05, 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(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1707e-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(0.0004, 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(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-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(4.5416e-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(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(0.0012, 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(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7841e-05, device='cuda:0', grad_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(6.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8558e-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(2.2224e-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(5.8587e-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(9.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-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(5.1985e-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(0.0003, 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(0.0004, 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(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(9.2053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(9.4734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3260e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3814e-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(0.0014, 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(4.4992e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_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(7.1391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-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(0.0006, 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(7.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1019e-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(1.8954e-05, 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(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4906e-05, device='cuda:0', grad_fn=<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(3.8998e-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(0.0006, 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(0.0005, 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(2.9400e-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(4.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-05, device='cuda:0', grad_fn=<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(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0001, 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(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(9.9574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-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(3.0732e-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(3.5123e-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(6.0646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9211e-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(2.9945e-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(7.7567e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0935e-05, 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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(6.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3919e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9094e-05, 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(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8133e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-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(0.0012, 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(0.0004, 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(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2368e-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(0.0004, 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(8.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7364e-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(2.8371e-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(0.0005, 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(4.2026e-05, 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(0.0006, 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(0.0003, 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(3.4729e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6415e-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(6.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<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(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1333e-05, 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(2.2224e-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(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-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(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5834e-05, device='cuda:0', grad_fn=<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(9.7488e-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(5.8716e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1141e-05, device='cuda:0', grad_fn=<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(3.5969e-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(4.8595e-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(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5904e-05, device='cuda:0', grad_fn=<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(4.8232e-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(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2767e-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(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-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(8.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9987e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-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(9.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5966e-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(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-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(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1843e-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
				 {'precision': {'support': 0.7091273821464393, 'agreement': 0.4619883040935672, 'direct_attack': 0.2, 'undercutter_attack': 0.35398230088495575, 'partial': 0.2638888888888889}, 'recall': {'support': 0.8643031784841075, 'agreement': 0.4030612244897959, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.20408163265306123, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7790633608815426, 'agreement': 0.4305177111716621, 'direct_attack': 0.16170212765957445, 'undercutter_attack': 0.2588996763754045, 'partial': 0.22352941176470587}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5966850828729282, 'recall': 0.5966850828729282, 'f1': 0.5966850828729282, 'support': None}, 'macro_avg': {'precision': 0.3977973752027702, 'recall': 0.3602075744723317, 'f1': 0.37074245757057794, 'support': None}, 'weighted_avg': {'precision': 0.5482441631786525, 'recall': 0.5966850828729282, 'f1': 0.5641877191889562, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6000e-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(0.0004, 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(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-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(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, 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(4.5386e-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(6.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5357e-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(8.4195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, 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(3.7423e-05, 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(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6275e-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(4.7324e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8935e-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(0.0001, 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(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5842e-05, device='cuda:0', grad_fn=<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(6.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-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(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6965e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-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(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-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(0.0003, 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(0.0001, 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(9.6183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-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(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8504e-05, device='cuda:0', grad_fn=<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(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-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(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-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(6.1704e-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(7.6234e-05, device='cuda:0', grad_fn=<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(4.6173e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, 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(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(9.6185e-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(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1918e-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(1.9409e-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(4.8353e-05, device='cuda:0', grad_fn=<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(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<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(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-05, 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(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<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(4.3542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5807e-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(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1254e-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(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8377e-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(3.0762e-05, device='cuda:0', grad_fn=<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(3.2367e-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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-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(0.0001, 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(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7899e-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(0.0008, 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(9.8645e-05, device='cuda:0', grad_fn=<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(0.0002, 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(8.9336e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8480e-05, device='cuda:0', grad_fn=<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(2.7735e-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(0.0005, 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(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9442e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-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(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<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(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-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(4.1541e-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(6.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7567e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5194e-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(0.0004, 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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-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(4.4508e-05, 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(5.8674e-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(0.0001, 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(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3070e-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(3.9331e-05, 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(2.2739e-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(0.0002, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8498e-05, 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(0.0003, 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(1.7683e-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(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-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(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(1.1748e-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(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8551e-05, 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(6.4369e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3822e-05, device='cuda:0', grad_fn=<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(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<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(0.0022, 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(2.1498e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, 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(6.4205e-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(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9242e-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(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(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5589e-05, 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(0.0001, 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(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(0.0005, 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(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1653e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7575e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-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(6.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, 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(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, 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(8.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7219e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9664e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7656e-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(8.1595e-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(4.5416e-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(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4831e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, 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(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(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4679e-05, device='cuda:0', grad_fn=<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(2.3042e-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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8087e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0272e-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(2.6463e-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(3.4002e-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(9.8597e-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(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7261e-05, 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(0.0004, 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(0.0002, 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(6.4097e-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(3.1156e-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(9.5701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-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(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-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(0.0004, 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(9.3559e-05, device='cuda:0', grad_fn=<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(8.5911e-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(8.1352e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3259e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, 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(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-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(6.6155e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7163e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6895e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8715e-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(1.5715e-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(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(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-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(0.0006, 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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7094188376753507, 'agreement': 0.45930232558139533, 'direct_attack': 0.20833333333333334, 'undercutter_attack': 0.36036036036036034, 'partial': 0.2676056338028169}, 'recall': {'support': 0.8644688644688645, 'agreement': 0.4030612244897959, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.20512820512820512, 'partial': 0.19387755102040816}, 'f1': {'support': 0.779306549257017, 'agreement': 0.4293478260869565, 'direct_attack': 0.1694915254237288, 'undercutter_attack': 0.261437908496732, 'partial': 0.22485207100591714}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 195, 'partial': 98}, 'micro_avg': {'precision': 0.5980662983425414, 'recall': 0.5980662983425414, 'f1': 0.5980662983425414, 'support': None}, 'macro_avg': {'precision': 0.40100409815065136, 'recall': 0.36187859759288327, 'f1': 0.3728871760540703, 'support': None}, 'weighted_avg': {'precision': 0.5502068873754687, 'recall': 0.5980662983425414, 'f1': 0.5657105983627797, 'support': None}}
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(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(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<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(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-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(3.5546e-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(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9673e-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(3.2488e-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(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-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(0.0005, 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(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(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8081e-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(9.2498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(3.4214e-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(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(0.0004, 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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, 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(0.0001, 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(4.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0003, 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(0.0003, 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(4.1571e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1579e-05, 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(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<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(9.0146e-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(9.8276e-05, device='cuda:0', grad_fn=<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(8.6721e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8578e-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(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5017e-05, 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(0.0001, 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(8.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-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(0.0004, 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(0.0002, 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(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(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, 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(0.0017, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6735e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0001, 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(0.0001, 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(0.0003, 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(2.2709e-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(0.0002, 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(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9985e-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(0.0006, 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(2.6615e-05, device='cuda:0', grad_fn=<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(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8880e-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(0.0003, 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(8.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4672e-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(0.0002, 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(5.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-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(7.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, 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(2.0983e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-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(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0002, device='cuda:0', grad_fn=<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(3.0581e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(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(0.0004, 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(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, 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(0.0003, 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(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, 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(4.8690e-05, 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(2.6827e-05, 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(0.0003, 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(3.1489e-05, device='cuda:0', grad_fn=<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(4.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7585e-05, 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(0.0003, 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(2.7674e-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(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1642e-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(0.0001, 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(3.5546e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6143e-05, device='cuda:0', grad_fn=<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(9.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.3708e-05, 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(0.0002, 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(1.7047e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5921e-05, device='cuda:0', grad_fn=<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(3.0975e-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(2.7311e-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(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2695e-05, 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(4.3600e-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(8.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9499e-05, device='cuda:0', grad_fn=<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(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5349e-05, 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(2.6039e-05, device='cuda:0', grad_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(5.2865e-05, 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(9.8001e-05, device='cuda:0', grad_fn=<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(8.5734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5318e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0082e-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(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-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(8.8839e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6083e-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(2.9430e-05, device='cuda:0', grad_fn=<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(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<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(0.0004, 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(3.9845e-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(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(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0060e-05, device='cuda:0', grad_fn=<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(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0190e-05, 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(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(6.3339e-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(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2110e-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(2.8250e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, 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(6.0306e-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(0.0004, 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(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-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(1.1203e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6677e-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(3.3729e-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(7.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-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(5.0836e-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(0.0005, 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(0.0008, 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(0.0005, 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(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(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9996e-05, device='cuda:0', grad_fn=<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(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.1467e-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(0.0002, 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(0.0002, 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(6.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-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(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, 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(8.6066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0602e-05, 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(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5352e-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(2.0620e-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(7.1452e-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(3.5879e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-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(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(7.1333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, 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(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, 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(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5255e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4742e-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(7.7235e-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(9.0763e-05, 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(1.7138e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7087087087087087, 'agreement': 0.45930232558139533, 'direct_attack': 0.20833333333333334, 'undercutter_attack': 0.36036036036036034, 'partial': 0.2714285714285714}, 'recall': {'support': 0.8655256723716381, 'agreement': 0.4030612244897959, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.20408163265306123, 'partial': 0.19387755102040816}, 'f1': {'support': 0.779306549257017, 'agreement': 0.4293478260869565, 'direct_attack': 0.1694915254237288, 'undercutter_attack': 0.26058631921824105, 'partial': 0.2261904761904762}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5980662983425414, 'recall': 0.5980662983425414, 'f1': 0.5980662983425414, 'support': None}, 'macro_avg': {'precision': 0.4016266598824738, 'recall': 0.36188064467840925, 'f1': 0.3729845392352839, 'support': None}, 'weighted_avg': {'precision': 0.5498233956042642, 'recall': 0.5980662983425414, 'f1': 0.5653282665732372, 'support': None}}
Loss: tensor(0.0004, 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(0.0003, 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(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, 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(6.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-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(2.3678e-05, 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(0.0004, 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(9.0228e-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(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-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(0.0002, 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(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4088e-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(8.9684e-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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4062e-05, device='cuda:0', grad_fn=<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(7.5115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-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(4.4508e-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(0.0002, 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(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.4806e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, 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(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(3.1489e-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(4.1299e-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(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-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(0.0003, 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(7.1764e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-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(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(1.4322e-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(5.9190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9525e-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(6.6458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<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(1.8137e-05, device='cuda:0', grad_fn=<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(7.1080e-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(5.4438e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-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(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, 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(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(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8555e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5013e-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(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, 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(0.0006, 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(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-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(0.0002, 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(6.8181e-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(2.2255e-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(3.5486e-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(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4937e-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(8.6008e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8799e-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(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8922e-05, 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(6.9755e-05, device='cuda:0', grad_fn=<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(8.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, 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(5.5213e-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(7.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-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(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3177e-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(5.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4094e-05, 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(5.4863e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6614e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2770e-05, device='cuda:0', grad_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(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(5.2193e-05, device='cuda:0', grad_fn=<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(4.5936e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0168e-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(5.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8514e-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(0.0001, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3791e-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(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8678e-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(5.2167e-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(2.1922e-05, 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(8.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4294e-05, device='cuda:0', grad_fn=<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(0.0018, 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(0.0001, 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(3.5002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-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(4.1540e-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(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, 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(0.0005, 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(9.1618e-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(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0293e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-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(5.4501e-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(2.8734e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-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(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-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(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9383e-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(6.0943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6103e-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(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(4.6264e-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(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(5.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9597e-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(0.0001, 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(0.0008, 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(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-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(7.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-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(0.0002, 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(3.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<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(0.0003, 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(5.2653e-05, 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(4.5356e-05, device='cuda:0', grad_fn=<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(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-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(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-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(3.9422e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2207e-05, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0388e-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(6.9814e-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(4.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4307e-05, 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(0.0002, 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(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1581e-05, device='cuda:0', grad_fn=<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(1.3444e-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(1.7592e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-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(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, 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(7.0513e-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(7.4875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<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(5.2680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6415e-05, device='cuda:0', grad_fn=<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(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-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(6.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5611e-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(8.7472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-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(2.1952e-05, device='cuda:0', grad_fn=<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(0.0002, 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(4.6657e-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(2.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-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(1.7047e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(7.7386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8504e-05, device='cuda:0', grad_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(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(2.8522e-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(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(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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1465e-05, device='cuda:0', grad_fn=<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(9.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3585e-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(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(3.0157e-05, device='cuda:0', grad_fn=<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(3.4083e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-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(8.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9209e-05, device='cuda:0', grad_fn=<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(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-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(0.0002, device='cuda:0', grad_fn=<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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-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(4.5386e-05, device='cuda:0', grad_fn=<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(2.5010e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8244e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9396e-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(2.7462e-05, device='cuda:0', grad_fn=<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(5.4378e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5669e-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>)
			------------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
				 {'precision': {'support': 0.7097097097097097, 'agreement': 0.45930232558139533, 'direct_attack': 0.20618556701030927, 'undercutter_attack': 0.36036036036036034, 'partial': 0.2753623188405797}, 'recall': {'support': 0.8656898656898657, 'agreement': 0.40512820512820513, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.20408163265306123, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7799779977997798, 'agreement': 0.4305177111716621, 'direct_attack': 0.16877637130801687, 'undercutter_attack': 0.26058631921824105, 'partial': 0.2275449101796407}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5987569060773481, 'recall': 0.5987569060773481, 'f1': 0.5987569060773481, 'support': None}, 'macro_avg': {'precision': 0.4021840563004709, 'recall': 0.36232687946973663, 'f1': 0.3734806619354681, 'support': None}, 'weighted_avg': {'precision': 0.5506203888115159, 'recall': 0.5987569060773481, 'f1': 0.5661297967016549, 'support': None}}
Loss: tensor(6.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(0.0003, 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(6.4656e-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(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-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(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969e-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(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-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(0.0002, 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(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-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(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-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(3.4850e-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(0.0001, 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(7.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0051e-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(0.0003, 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(5.1622e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, 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(0.0007, 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(7.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<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(2.8068e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-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(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5845e-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(0.0002, 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(3.9967e-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(0.0004, 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(7.6102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<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(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3293e-05, 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(3.7983e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, 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(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6409e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6720e-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(5.3621e-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(0.0006, 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(6.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, 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(0.0001, 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(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-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(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-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(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, 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(0.0003, 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(0.0001, 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(2.5979e-05, device='cuda:0', grad_fn=<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(0.0004, 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(8.2817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, 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(7.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9451e-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(9.5283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8020e-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(3.1550e-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(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6410e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, 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(0.0002, 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(7.9587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2117e-05, device='cuda:0', grad_fn=<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(2.2345e-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(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(6.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7274e-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(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-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(0.0007, 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(0.0016, 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(3.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6746e-05, device='cuda:0', grad_fn=<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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-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(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6012e-05, device='cuda:0', grad_fn=<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(1.9832e-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(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-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(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-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(0.0005, 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(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(0.0005, 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(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8676e-05, 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(0.0003, 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(6.1082e-05, device='cuda:0', grad_fn=<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(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9433e-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(0.0002, 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(0.0002, 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(6.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-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(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-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(2.7614e-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(2.9219e-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(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-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(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7121e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2565e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1342e-05, 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(6.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6477e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9162e-05, device='cuda:0', grad_fn=<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(7.4023e-05, device='cuda:0', grad_fn=<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(1.8652e-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(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8584e-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(4.0198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-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(7.7171e-05, device='cuda:0', grad_fn=<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(4.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4488e-05, 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(2.8861e-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(9.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-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(6.4019e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3326e-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(3.2700e-05, device='cuda:0', grad_fn=<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(3.9967e-05, 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(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5927e-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(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(0.0003, 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(4.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8282e-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(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0617e-05, device='cuda:0', grad_fn=<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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-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(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, 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(0.0005, 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(4.0723e-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(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2567e-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(4.2298e-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(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6989e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-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(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6209e-05, device='cuda:0', grad_fn=<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(4.4296e-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(0.0001, 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(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2074e-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(2.8371e-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(8.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, 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(1.4958e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-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(9.2461e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7668e-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(8.9616e-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(2.6554e-05, 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(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1123e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5176e-05, device='cuda:0', grad_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(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(8.7591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7667e-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(8.3471e-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(7.0240e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<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(4.1264e-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(6.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3327e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4170e-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(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2332e-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(3.0187e-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(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-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(0.0001, 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(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0455e-05, device='cuda:0', grad_fn=<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(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-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(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(0.0001, 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(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8329e-05, 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(0.0003, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7171e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0840e-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(8.5756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5247e-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(3.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, 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(6.1664e-05, 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(2.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4169e-05, device='cuda:0', grad_fn=<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>)
			------------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
				 {'precision': {'support': 0.7111334002006018, 'agreement': 0.4624277456647399, 'direct_attack': 0.20618556701030927, 'undercutter_attack': 0.35714285714285715, 'partial': 0.2753623188405797}, 'recall': {'support': 0.8667481662591687, 'agreement': 0.40816326530612246, 'direct_attack': 0.14285714285714285, 'undercutter_attack': 0.20408163265306123, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7812672176308539, 'agreement': 0.4336043360433604, 'direct_attack': 0.16877637130801687, 'undercutter_attack': 0.2597402597402597, 'partial': 0.2275449101796407}, 'support': {'support': 818, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5994475138121547, 'recall': 0.5994475138121547, 'f1': 0.5994475138121547, 'support': None}, 'macro_avg': {'precision': 0.4024503777718175, 'recall': 0.36314555161918066, 'f1': 0.3741866189804263, 'support': None}, 'weighted_avg': {'precision': 0.5512392583854981, 'recall': 0.5994475138121547, 'f1': 0.5669200400389193, 'support': None}}
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3199e-05, device='cuda:0', grad_fn=<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(3.8728e-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(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0736e-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(9.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8663e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(9.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6120e-05, device='cuda:0', grad_fn=<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(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, 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(6.2112e-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(5.9554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-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(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(9.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<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(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0837e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6001e-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(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6074e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-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(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4002e-05, 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(0.0001, 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(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<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(5.0079e-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(5.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7524e-05, device='cuda:0', grad_fn=<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(7.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6466e-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(2.8193e-05, device='cuda:0', grad_fn=<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(1.9863e-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(8.2402e-05, 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(4.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1596e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3782e-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(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(4.1026e-05, device='cuda:0', grad_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(4.0208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286e-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(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(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-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(3.6364e-05, device='cuda:0', grad_fn=<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(0.0001, 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(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-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(0.0006, 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(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(2.8734e-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(4.2722e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0288e-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(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, 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(0.0004, 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(4.2025e-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(9.7685e-05, 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(0.0001, 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(2.9309e-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(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(9.7968e-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(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4309e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6987e-05, 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(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, 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(3.1459e-05, device='cuda:0', grad_fn=<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(7.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-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(6.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-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>)
Loss: tensor(4.1935e-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(9.9920e-06, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3583e-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(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0154e-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(3.6727e-05, device='cuda:0', grad_fn=<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(1.8954e-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(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0452e-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(9.8087e-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(1.2354e-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(3.6074e-05, 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(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, 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(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1957e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1785e-05, device='cuda:0', grad_fn=<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(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8159e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7538e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5577e-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(5.8040e-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(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(8.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4186e-05, 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(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0262e-05, device='cuda:0', grad_fn=<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(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-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(5.2500e-05, 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(6.0056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(5.4893e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-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(1.8773e-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(0.0001, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8325e-05, 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(4.3176e-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(5.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9799e-05, 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(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8303e-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(0.0003, 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(2.4465e-05, device='cuda:0', grad_fn=<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(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-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(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, 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(3.2610e-05, 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(0.0002, device='cuda:0', grad_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(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2430e-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(7.1711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, 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(4.3054e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-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(1.6774e-05, 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(6.6578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2645e-05, 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(4.2540e-05, device='cuda:0', grad_fn=<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(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8225e-05, 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(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(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-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(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(1.5291e-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(5.2077e-05, 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(2.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-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(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7519e-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(3.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7059e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, 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(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-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(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4552e-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(0.0001, 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(4.8867e-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(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-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(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9720e-05, 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(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6286e-05, 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(0.0003, 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(5.2985e-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(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2636e-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(1.4897e-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(1.5079e-05, 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(0.0003, 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(8.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4209e-05, device='cuda:0', grad_fn=<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(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2063e-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(0.0002, 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(2.9097e-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(3.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-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(6.0849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-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(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9251e-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(8.0506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-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(0.0001, 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(2.0105e-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(2.6796e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-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(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(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-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(0.0001, 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(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5401e-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(6.0601e-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
				 {'precision': {'support': 0.7105788423153693, 'agreement': 0.45977011494252873, 'direct_attack': 0.20212765957446807, 'undercutter_attack': 0.3486238532110092, 'partial': 0.2753623188405797}, 'recall': {'support': 0.8693528693528694, 'agreement': 0.41025641025641024, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.19387755102040816, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7819879187259747, 'agreement': 0.4336043360433604, 'direct_attack': 0.16239316239316237, 'undercutter_attack': 0.24918032786885247, 'partial': 0.2275449101796407}, 'support': {'support': 819, 'agreement': 195, 'direct_attack': 140, 'undercutter_attack': 196, 'partial': 98}, 'micro_avg': {'precision': 0.5994475138121547, 'recall': 0.5994475138121547, 'f1': 0.5994475138121547, 'support': None}, 'macro_avg': {'precision': 0.39929255777679096, 'recall': 0.36061573347287623, 'f1': 0.3709421310421982, 'support': None}, 'weighted_avg': {'precision': 0.5491939910816579, 'recall': 0.5994475138121547, 'f1': 0.5655212287016375, 'support': None}}
Loss: tensor(0.0002, 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(4.1087e-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(1.5654e-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(0.0003, 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(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-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(8.9581e-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(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, 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(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2426e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6985e-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(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1298e-05, device='cuda:0', grad_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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6344e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8154e-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(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0603e-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(9.2363e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-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(2.7038e-05, 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(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-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(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7112e-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(0.0002, 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(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6303e-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(0.0001, device='cuda:0', grad_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(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8867e-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(3.0460e-05, 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(1.2929e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<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(4.7976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0888e-05, device='cuda:0', grad_fn=<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(1.6835e-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(2.0983e-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(3.4917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-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(0.0001, 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(2.7886e-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(0.0003, 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(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0364e-05, 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(0.0002, 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(5.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-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(8.5375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_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(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6396e-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(6.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-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(1.6714e-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(0.0003, 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(1.9045e-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(2.5101e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5416e-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(0.0001, 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(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8187e-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(0.0005, 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(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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4290e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5196e-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(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, 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(4.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0975e-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(5.5951e-05, 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(0.0002, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9595e-05, device='cuda:0', grad_fn=<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(1.9136e-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(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<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(4.9020e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1667e-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(0.0002, 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(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-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(0.0002, 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(0.0002, 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(6.3615e-05, device='cuda:0', grad_fn=<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(3.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-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(8.2113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-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(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-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(5.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5864e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6361e-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(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9623e-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(6.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1241e-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(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, 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(9.7730e-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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3210e-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(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203e-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(7.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4498e-05, device='cuda:0', grad_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(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-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(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0838e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8494e-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(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, 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(0.0002, 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(6.0310e-05, device='cuda:0', grad_fn=<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(5.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_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(9.2110e-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(0.0002, 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(0.0001, 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(3.8532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4711e-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(0.0003, 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(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8534e-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(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-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(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(0.0002, 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(2.8250e-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(4.4448e-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(2.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(5.7178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, 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(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7111e-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(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8704e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-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(4.0087e-05, device='cuda:0', grad_fn=<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(0.0001, 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(2.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0853e-05, 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(0.0002, 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(4.3480e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-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(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-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(6.8772e-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(9.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8089e-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(7.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9709e-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(0.0004, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6436e-05, device='cuda:0', grad_fn=<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(0.0002, 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(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-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(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<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(4.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0313e-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(0.0001, 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(8.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-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(3.7651e-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(2.5098e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9715e-05, 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(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9906e-05, device='cuda:0', grad_fn=<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(6.1006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-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(0.0003, 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(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, 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(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-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(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-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(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<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(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-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
				 {'precision': {'support': 0.7091633466135459, 'agreement': 0.45977011494252873, 'direct_attack': 0.20212765957446807, 'undercutter_attack': 0.34579439252336447, 'partial': 0.2753623188405797}, 'recall': {'support': 0.8693528693528694, 'agreement': 0.40816326530612246, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.18974358974358974, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7811300054854636, 'agreement': 0.4324324324324324, 'direct_attack': 0.16239316239316237, 'undercutter_attack': 0.24503311258278146, 'partial': 0.2275449101796407}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 195, 'partial': 98}, 'micro_avg': {'precision': 0.5987569060773481, 'recall': 0.5987569060773481, 'f1': 0.5987569060773481, 'support': None}, 'macro_avg': {'precision': 0.39844356649889734, 'recall': 0.3593703122274551, 'f1': 0.3697067246146961, 'support': None}, 'weighted_avg': {'precision': 0.5480890949821051, 'recall': 0.5987569060773481, 'f1': 0.5644462238505811, 'support': None}}
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(8.7809e-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(0.0002, 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(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4185e-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(4.0137e-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(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4452e-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(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6238e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7425e-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(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, 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(4.7266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-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(8.9407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2028e-05, device='cuda:0', grad_fn=<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(3.2428e-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(1.9439e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1378e-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(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0444e-05, device='cuda:0', grad_fn=<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(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3112e-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(5.8283e-05, 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(0.0002, 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(0.0003, 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(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2493e-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(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3408e-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(2.0650e-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(3.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2905e-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(0.0002, 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(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-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(5.7994e-05, device='cuda:0', grad_fn=<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(3.7271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5990e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0819e-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(5.6421e-05, 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(8.4116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1221e-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(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4091e-05, 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(1.0598e-05, 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(3.4998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8463e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9323e-05, device='cuda:0', grad_fn=<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(5.9204e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-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(4.8805e-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(0.0002, 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(7.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2681e-05, device='cuda:0', grad_fn=<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(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-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(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4578e-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(0.0002, 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(6.9358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0882e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7113e-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(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(8.2818e-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(7.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-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(0.0002, 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(8.1292e-05, 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(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6874e-05, 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(1.8046e-05, 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(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1170e-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(0.0001, 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(2.3436e-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(4.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9251e-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(0.0002, 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(9.1830e-05, device='cuda:0', grad_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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6591e-05, 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(2.6796e-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(0.0002, 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(8.3515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8275e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6341e-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(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2252e-05, device='cuda:0', grad_fn=<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(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0884e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-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(7.6173e-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(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(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2881e-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(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3933e-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(6.8562e-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(0.0003, 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(2.1800e-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(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9052e-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(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1229e-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(0.0004, 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(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-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(0.0003, 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(5.0188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, 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(5.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4825e-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(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-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(0.0003, 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(9.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8347e-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(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(5.0905e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, 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(3.1436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, 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(6.9107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0854e-05, device='cuda:0', grad_fn=<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(1.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7550e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-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(0.0002, 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(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-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(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5325e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-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(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8041e-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(4.6142e-05, device='cuda:0', grad_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(0.0001, 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(1.3202e-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(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, 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(2.4556e-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(2.5161e-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(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, 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(3.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1896e-05, 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(1.3504e-05, device='cuda:0', grad_fn=<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(1.6895e-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(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8680e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-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(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, 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(0.0001, 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(4.4932e-05, device='cuda:0', grad_fn=<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(9.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3217e-05, device='cuda:0', grad_fn=<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(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1591e-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(7.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4340e-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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-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(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, 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(0.0003, 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(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1697e-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(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(8.2160e-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(9.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-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(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0688e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-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(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2066e-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(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9614e-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(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9341e-05, 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(4.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9294e-05, 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(3.8919e-05, 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(9.9418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2899e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-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(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(1.2414e-05, 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(8.4381e-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(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(2.9249e-05, 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(1.7622e-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(7.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3358e-05, device='cuda:0', grad_fn=<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(3.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7353e-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(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-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
				 {'precision': {'support': 0.709452736318408, 'agreement': 0.45977011494252873, 'direct_attack': 0.2, 'undercutter_attack': 0.3523809523809524, 'partial': 0.2753623188405797}, 'recall': {'support': 0.8705738705738706, 'agreement': 0.40816326530612246, 'direct_attack': 0.1357142857142857, 'undercutter_attack': 0.18974358974358974, 'partial': 0.19387755102040816}, 'f1': {'support': 0.7817982456140351, 'agreement': 0.4324324324324324, 'direct_attack': 0.16170212765957445, 'undercutter_attack': 0.24666666666666662, 'partial': 0.2275449101796407}, 'support': {'support': 819, 'agreement': 196, 'direct_attack': 140, 'undercutter_attack': 195, 'partial': 98}, 'micro_avg': {'precision': 0.5994475138121547, 'recall': 0.5994475138121547, 'f1': 0.5994475138121547, 'support': None}, 'macro_avg': {'precision': 0.39939322449649384, 'recall': 0.3596145124716553, 'f1': 0.3700288765104699, 'support': None}, 'weighted_avg': {'precision': 0.548934065286032, 'recall': 0.5994475138121547, 'f1': 0.5649773611772078, '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
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.08463251670378619, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.9827586206896551, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.15584415584415584, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.08431952662721894, 'recall': 0.08431952662721894, 'f1': 0.08431952662721894, 'support': None}, 'macro_avg': {'precision': 0.016926503340757237, 'recall': 0.19655172413793104, 'f1': 0.03116883116883117, 'support': None}, 'weighted_avg': {'precision': 0.007261369776360354, 'recall': 0.08431952662721894, 'f1': 0.01337124414047491, 'support': None}}
Loss: tensor(2.9694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, 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
				 {'precision': {'support': 0.8673913043478261, 'agreement': 0.3240223463687151, 'direct_attack': 0.149812734082397, 'undercutter_attack': 0.0935672514619883, 'partial': 0.10416666666666667}, 'recall': {'support': 0.4634146341463415, 'agreement': 0.7073170731707317, 'direct_attack': 0.3448275862068966, 'undercutter_attack': 0.1322314049586777, 'partial': 0.1111111111111111}, 'f1': {'support': 0.6040878122634369, 'agreement': 0.4444444444444444, 'direct_attack': 0.20887728459530025, 'undercutter_attack': 0.10958904109589042, 'partial': 0.1075268817204301}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.4297337278106509, 'recall': 0.4297337278106509, 'f1': 0.42973372781065095, 'support': None}, 'macro_avg': {'precision': 0.3077920605855186, 'recall': 0.3517803619187517, 'f1': 0.2949050928239004, 'support': None}, 'weighted_avg': {'precision': 0.6198509559381702, 'recall': 0.4297337278106509, 'f1': 0.4735029242516304, 'support': None}}
Loss: tensor(3.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, 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
				 {'precision': {'support': 0.9054794520547945, 'agreement': 0.6103896103896104, 'direct_attack': 0.14563106796116504, 'undercutter_attack': 0.23357664233576642, 'partial': 0.2087912087912088}, 'recall': {'support': 0.7677119628339141, 'agreement': 0.573170731707317, 'direct_attack': 0.12931034482758622, 'undercutter_attack': 0.5289256198347108, 'partial': 0.2111111111111111}, 'f1': {'support': 0.830923947203017, 'agreement': 0.5911949685534591, 'direct_attack': 0.136986301369863, 'undercutter_attack': 0.3240506329113924, 'partial': 0.2099447513812155}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6309171597633136, 'recall': 0.6309171597633136, 'f1': 0.6309171597633136, 'support': None}, 'macro_avg': {'precision': 0.420773596306509, 'recall': 0.44204595406292785, 'f1': 0.4186201202837895, 'support': None}, 'weighted_avg': {'precision': 0.6979799487576965, 'recall': 0.6309171597633136, 'f1': 0.6556043332470834, 'support': None}}
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, 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(1.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8988, 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
				 {'precision': {'support': 0.9076175040518638, 'agreement': 0.4142857142857143, 'direct_attack': 0.19066147859922178, 'undercutter_attack': 0.21637426900584794, 'partial': 0.25925925925925924}, 'recall': {'support': 0.6504065040650406, 'agreement': 0.7073170731707317, 'direct_attack': 0.4224137931034483, 'undercutter_attack': 0.30578512396694213, 'partial': 0.07777777777777778}, 'f1': {'support': 0.7577807848443842, 'agreement': 0.5225225225225225, 'direct_attack': 0.2627345844504021, 'undercutter_attack': 0.25342465753424653, 'partial': 0.11965811965811966}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.5687869822485208, 'recall': 0.5687869822485208, 'f1': 0.5687869822485208, 'support': None}, 'macro_avg': {'precision': 0.3976396450403814, 'recall': 0.43274005441678814, 'f1': 0.383224133801935, 'support': None}, 'weighted_avg': {'precision': 0.681237336931999, 'recall': 0.5687869822485208, 'f1': 0.5991522008667379, 'support': None}}
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, 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(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9189, 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(0.6097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, 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(0.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, 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(0.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, 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(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, 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(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, 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(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, 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(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4511, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4692, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, 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(0.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, 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(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, 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(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, 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(0.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, 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(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4984, 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(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9925, 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(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, 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(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, 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(0.0753, 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(0.0018, 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(1.2845, 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(0.0117, 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(0.0024, 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(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1812, 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(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, 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(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7125, 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(0.0325, 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(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, 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(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, 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(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, 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(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, 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(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, 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(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2106, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, 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(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, 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(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1391, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, 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(0.0083, 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(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.3624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, 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(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, 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(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, 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(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, 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(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0920, 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(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, 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(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, 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(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, 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(0.9124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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
				 {'precision': {'support': 0.8907894736842106, 'agreement': 0.43902439024390244, 'direct_attack': 0.21686746987951808, 'undercutter_attack': 0.30808080808080807, 'partial': 0.2153846153846154}, 'recall': {'support': 0.7862950058072009, 'agreement': 0.6585365853658537, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5041322314049587, 'partial': 0.15555555555555556}, 'f1': {'support': 0.8352868599629858, 'agreement': 0.526829268292683, 'direct_attack': 0.18090452261306533, 'undercutter_attack': 0.3824451410658307, 'partial': 0.18064516129032257}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6494082840236687, 'recall': 0.6494082840236687, 'f1': 0.6494082840236687, 'support': None}, 'macro_avg': {'precision': 0.41402935145461095, 'recall': 0.45193835838533447, 'f1': 0.42122219064497746, 'support': None}, 'weighted_avg': {'precision': 0.6810567725669545, 'recall': 0.6494082840236687, 'f1': 0.6576189627487729, 'support': None}}
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, 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(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, 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(0.0058, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, 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(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, 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(0.0037, 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(0.0085, 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(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, 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(0.0201, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, 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(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.2897, 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(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2627, 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(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, 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(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, 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(0.0469, 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(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, 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(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, 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(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, 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(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, 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(0.0074, 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(0.0171, 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(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, 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(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6310, 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(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, 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(0.0024, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, 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(0.0150, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, 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(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, 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(0.0036, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, 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(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0193, 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(0.0041, 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(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, 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(0.0052, 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(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0296, 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(0.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2633, 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(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, 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(0.0025, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4366, 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(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, 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(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, 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(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, 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(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2204, 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(1.9206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, 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(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, 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(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, 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(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0012, 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(0.1266, 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(0.0011, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, 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(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, 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(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, 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(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0971, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0031, 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(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3125, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9566, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, 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(0.0265, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, 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(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7350, 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(0.0133, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9947, 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(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, 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(0.7347, 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(0.0049, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, 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(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4641, 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(0.1081, 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(0.0044, 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(0.0043, 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(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, 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(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, 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(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9303, 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(0.0053, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1944, 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(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, 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(0.0627, 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(0.0097, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, 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(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, 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(0.0005, 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(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, 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(0.0056, 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(0.0060, 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(0.0471, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, 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(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, 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(0.0134, 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(0.0051, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, 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(0.4536, 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(0.0911, 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(0.0027, 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(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, 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(0.0053, 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(0.0103, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4161, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0040, 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(0.1267, 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(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7388, 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(0.3611, 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(0.0032, 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(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, 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(0.0442, 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(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, 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(0.0012, 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(0.0049, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, 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(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_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(0.0467, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, 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(0.0025, 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(0.0061, 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(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, 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(0.0037, 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(0.0079, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, 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(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, 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(0.0113, 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(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0019, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, 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
				 {'precision': {'support': 0.9005763688760807, 'agreement': 0.4246031746031746, 'direct_attack': 0.20187793427230047, 'undercutter_attack': 0.27358490566037735, 'partial': 0.21839080459770116}, 'recall': {'support': 0.7259001161440186, 'agreement': 0.6524390243902439, 'direct_attack': 0.3706896551724138, 'undercutter_attack': 0.2396694214876033, 'partial': 0.2111111111111111}, 'f1': {'support': 0.8038585209003215, 'agreement': 0.514423076923077, 'direct_attack': 0.2613981762917933, 'undercutter_attack': 0.2555066079295154, 'partial': 0.2146892655367232}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6087278106508875, 'recall': 0.6087278106508875, 'f1': 0.6087278106508875, 'support': None}, 'macro_avg': {'precision': 0.40380663760192687, 'recall': 0.4399618656610781, 'f1': 0.4099751295162862, 'support': None}, 'weighted_avg': {'precision': 0.6813668347718281, 'recall': 0.6087278106508875, 'f1': 0.6339113114039837, 'support': None}}
Loss: tensor(0.0059, 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(0.0015, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, 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(0.1281, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, 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(0.1464, 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(0.0102, 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(0.0005, 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(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, 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(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, 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(0.0005, 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(0.0172, 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(0.0026, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, 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(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0026, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, 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(0.0016, 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(0.0049, 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(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, 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(0.6179, 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(0.0339, device='cuda:0', grad_fn=<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(0.0091, 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(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0051, 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(0.0297, 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(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, 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(0.0048, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, 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(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, 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(0.0125, 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(0.1205, 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(0.0003, 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(0.0002, 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(0.1551, 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(0.0016, 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(0.0014, 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(0.0137, 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(0.0092, 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(0.0078, 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(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, 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(0.0022, 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(0.0017, 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(0.0003, 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(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0021, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0075, device='cuda:0', grad_fn=<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(0.0013, 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(0.0009, 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(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, 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(0.0016, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, 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(0.0143, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, 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(0.0047, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, 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(0.0028, 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(0.0024, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, 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(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, 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(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, 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(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, 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(0.0217, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, 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(0.0040, 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(0.0127, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, 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(0.0017, 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(0.0072, 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(0.0103, 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(0.0063, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, 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(0.0032, 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(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0082, 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(0.0055, 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(0.0050, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, 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(0.0064, 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(0.0009, 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(0.0135, 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(1.8053, 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(0.0026, 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(0.0009, 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(0.0262, 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(0.0020, 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(0.0003, 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(0.0029, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, 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(0.0028, 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(0.0034, 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(0.0051, 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(0.0033, 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(0.0045, 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(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, 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(0.0025, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, 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(0.0123, 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(0.0004, 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(0.0025, 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(0.0029, 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(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, 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(0.0378, 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(0.0005, 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(0.0008, 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(0.0132, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, 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(0.0045, 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(0.0078, 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(0.0017, 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(0.0008, 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(0.0082, 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(0.0194, 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(0.0058, 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(0.0005, 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(0.0029, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, 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(0.0022, 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(0.0017, 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(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, 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(0.0022, 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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, 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(0.0002, 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(0.0122, 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(0.0056, 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(0.0216, 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(0.0334, 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(0.0016, 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(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, 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(0.0008, 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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, 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(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, 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(0.0045, 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(1.0908, 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(0.0002, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, 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(0.0015, 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(0.0034, 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(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, 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(0.0023, 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(0.0030, 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(0.0002, 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(0.0025, 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(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, 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(0.0111, 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(0.0015, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, 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(0.0144, 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(0.0051, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9672, 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(0.0024, 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(0.0009, 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(0.0051, 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(0.0010, 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(0.0273, 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(0.0009, 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(0.0006, 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(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(0.0125, 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(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2922, 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(0.0512, 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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, 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(0.0023, 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(0.0108, 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(0.0191, 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(0.0005, 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(0.0023, 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(0.0010, 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(0.0009, 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(0.0002, 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(0.0014, 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(0.0003, 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(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, 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(0.0005, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, 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(0.0035, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, 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(0.0019, 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(0.0002, 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(0.0101, 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(0.0092, 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(0.0004, 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(0.0065, 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(0.0045, 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(0.0037, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4354, 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(0.0026, 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(0.0014, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, 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(0.0013, 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(0.0273, 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(0.0025, 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(0.0004, 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(0.0076, 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(0.1924, 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(0.0003, 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(0.0034, 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(0.0014, 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(0.0159, 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(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0068, 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(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, 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(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8929, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, 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(0.0149, device='cuda:0', grad_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(0.0007, 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(0.0055, 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(0.0003, 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(0.0004, 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(0.0031, 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(0.0292, 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(0.0082, 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(0.0041, 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(0.0334, 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(0.0036, 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(0.1478, 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(0.0005, 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(0.0028, 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(0.0087, 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(0.0117, 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(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, 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(0.0039, 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(0.3396, 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(0.0043, 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(0.0061, 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(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, 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(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, 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
				 {'precision': {'support': 0.9126853377265239, 'agreement': 0.48058252427184467, 'direct_attack': 0.23684210526315788, 'undercutter_attack': 0.20465116279069767, 'partial': 0.24242424242424243}, 'recall': {'support': 0.6434378629500581, 'agreement': 0.6036585365853658, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.7272727272727273, 'partial': 0.08888888888888889}, 'f1': {'support': 0.7547683923705722, 'agreement': 0.5351351351351352, 'direct_attack': 0.1875, 'undercutter_attack': 0.3194192377495462, 'partial': 0.13008130081300812}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.5673076923076923, 'recall': 0.5673076923076923, 'f1': 0.5673076923076923, 'support': None}, 'macro_avg': {'precision': 0.4154370744952933, 'recall': 0.4436860858980287, 'f1': 0.38538081321365236, 'support': None}, 'weighted_avg': {'precision': 0.6942990136756674, 'recall': 0.5673076923076923, 'f1': 0.5989088704394163, 'support': None}}
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3642, 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(0.0314, 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(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, 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(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, 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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, 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(0.0034, 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(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3276, 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(0.0029, 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(0.0075, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<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(0.0031, 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(0.0059, 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(0.0037, 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(0.0019, device='cuda:0', grad_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(0.0025, 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(0.0050, 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(0.0042, 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(0.0003, 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(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(0.0016, 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(0.0024, 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(0.0004, 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(0.0004, 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(0.0003, 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(0.0258, 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(0.0009, 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(0.0017, 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(0.0028, 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(0.0014, 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(0.0014, 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(0.0051, 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(0.0050, 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(0.0064, 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(0.0017, 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(0.0056, 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(0.0034, 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(0.0065, 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(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, 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(0.0002, 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(0.0029, 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(0.0027, 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(0.0004, 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(0.0026, 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(0.0003, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, 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(0.0002, 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(0.0045, 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(0.0035, 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(0.0018, 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(0.0004, 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(0.0026, 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(0.0002, 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(0.0007, 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(0.0067, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, 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(0.0036, 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(0.0001, 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(0.0098, 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(0.0005, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, 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(0.0002, 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(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<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(0.0032, 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(0.0043, 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(0.0018, 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(0.0036, 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(0.0027, 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(0.0069, 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(0.0018, 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(0.0005, 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(0.0003, 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(0.0030, 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(0.0015, 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(0.0077, 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(0.0062, 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(0.0020, 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(0.0049, 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(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0049, 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(0.0099, 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(0.0010, 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(0.0009, 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(0.0023, 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(0.0001, 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(0.0011, 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(0.0027, 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(0.0018, 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(0.0020, 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(0.0053, 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(0.0038, 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(0.0026, 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(0.0023, 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(0.0064, 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(0.0060, 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(0.0046, 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(0.0029, 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(0.0005, 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(0.0004, 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(0.0012, 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(0.0001, 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(0.0027, 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(0.0028, 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(0.0007, 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(0.0002, 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(0.0013, 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(0.0029, 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(0.0018, 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(0.0053, 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(0.0003, 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(0.0020, 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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, 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(0.0028, 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(0.0004, 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(0.0014, 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(0.0106, 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(0.0308, 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(0.0038, 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(9.8216e-05, device='cuda:0', grad_fn=<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(0.0084, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0002, 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(0.0001, 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(0.0033, 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(0.0005, 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(0.0130, 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(0.0032, 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(0.0335, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, 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(0.0110, 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(0.0002, 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(0.0124, device='cuda:0', grad_fn=<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(0.0184, 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(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, 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(0.0005, 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(0.0005, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, 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(0.0041, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(0.0019, 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(0.0001, 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(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0056, 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(0.0008, 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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, 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(0.0010, 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(0.0032, 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(0.0080, 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(0.0015, 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(0.0002, 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(0.0010, 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(0.0002, 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(0.0049, 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(0.0026, 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(0.0003, 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(0.0002, 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(0.0019, 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(0.0035, 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(0.0044, 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(0.0172, 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(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, 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(0.0008, 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(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0018, 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(0.0005, 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(0.0032, 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(0.0016, 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(0.0027, 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(0.0012, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0043, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, 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(0.0082, 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(0.0011, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, 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(0.0001, 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(0.0020, 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(0.0001, 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(0.0015, 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(0.0243, 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(0.0193, device='cuda:0', grad_fn=<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(0.0013, 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(0.0007, 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(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0022, 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(0.0002, 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(0.0051, 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(0.0076, 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(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(0.0014, 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(0.0001, 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(0.0038, 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(0.0023, 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(0.0010, 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(0.0250, 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(0.0057, 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(9.1071e-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(0.0031, 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(0.0013, 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(0.0002, 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(0.0016, 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(0.0009, 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(0.0002, 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(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, 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(0.3485, 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(0.0002, 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(0.0073, 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(0.0004, 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(0.0047, 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(0.0047, device='cuda:0', grad_fn=<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(0.0023, 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(0.0026, 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(0.0064, 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(0.0026, 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(0.0004, 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(0.0012, 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(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, 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(0.0036, 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(0.0155, 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(0.0027, 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(0.0004, 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(0.0017, 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(0.0036, 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(0.0022, 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(0.0036, 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(0.0014, 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(0.0020, 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(0.0046, 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(0.0008, 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(0.0044, 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(0.0022, 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(0.0015, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2255, 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(0.0010, 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(0.0038, 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(0.0020, 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(0.0122, 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(0.0027, device='cuda:0', grad_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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3705e-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(0.0018, 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(0.0002, 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(0.0007, 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(0.0070, 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(0.0023, 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(0.0008, 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(0.0063, 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(0.0003, 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(0.0136, device='cuda:0', grad_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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6248e-05, 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(0.0095, 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(0.0011, 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(0.0034, 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(0.0008, 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(0.0006, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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
				 {'precision': {'support': 0.8917456021650879, 'agreement': 0.4646017699115044, 'direct_attack': 0.20689655172413793, 'undercutter_attack': 0.2620967741935484, 'partial': 0.23076923076923078}, 'recall': {'support': 0.7653890824622532, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.13333333333333333}, 'f1': {'support': 0.82375, 'agreement': 0.5384615384615384, 'direct_attack': 0.17733990147783252, 'undercutter_attack': 0.3523035230352304, 'partial': 0.16901408450704225}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6353550295857988, 'recall': 0.6353550295857988, 'f1': 0.6353550295857988, 'support': None}, 'macro_avg': {'precision': 0.41122198575270186, 'recall': 0.4462657629344685, 'f1': 0.41217380949632876, 'support': None}, 'weighted_avg': {'precision': 0.680821445396655, 'recall': 0.6353550295857988, 'f1': 0.6479052254230898, 'support': None}}
Loss: tensor(0.0021, 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(0.0035, 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(0.0028, 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(0.0001, 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(0.0023, 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(0.0036, 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(0.0012, 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(0.0010, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9093e-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(0.0041, device='cuda:0', grad_fn=<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(0.0108, 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(0.0014, 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(0.0013, 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(0.0021, 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(0.0026, 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(0.0014, 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(0.0009, 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(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, 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(0.0002, 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(0.0027, 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(0.0004, 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(0.0017, 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(0.0052, 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(0.0009, 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(0.0006, 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(0.0011, 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(0.0003, 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(0.0014, 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(0.0007, 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(0.0019, 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(0.0021, 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(0.0019, 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(0.0013, 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(0.0011, 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(0.0094, 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(0.0029, 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(0.0043, 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(0.0041, 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(0.0003, 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(0.0012, 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(0.0031, 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(0.0010, 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(0.0022, 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(0.0048, 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(0.0014, 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(0.0001, 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(0.0014, 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(0.0015, 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(0.0048, 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(0.0056, 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(9.7520e-05, 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(0.0029, 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(0.0033, 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(0.0016, 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(0.0001, 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(0.0048, 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(0.0069, 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(0.0014, 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(0.0019, 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(0.0013, 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(0.0011, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7671e-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(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(0.0040, 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(0.0001, 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(0.0028, 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(0.0036, 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(0.0009, 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(0.0010, 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(0.0015, 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(0.0031, 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(8.0597e-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(0.0015, 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(0.0213, device='cuda:0', grad_fn=<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(0.0025, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, 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(0.0003, 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(0.0001, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, 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(0.0024, 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(0.0002, 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(0.0021, 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(0.0017, 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(0.0002, 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(0.0019, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, 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(0.0006, 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(0.0009, 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(0.0014, 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(0.0009, 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(0.0027, 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(0.0036, 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(0.0007, 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(0.0023, 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(0.0021, 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(0.0001, device='cuda:0', grad_fn=<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(0.0049, 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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5804e-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(0.0027, 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(0.0004, 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(0.0007, 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(0.0032, device='cuda:0', grad_fn=<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(0.0037, 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(0.0021, 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(0.0004, 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(0.0034, 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(0.0026, 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(0.0009, 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(0.0020, 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(0.0038, 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(0.0056, 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(0.0009, 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(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(0.0002, 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(0.0011, 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(0.0023, 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(0.0001, 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(0.0024, 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(0.0024, 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(0.0017, 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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0004, 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(0.0015, 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(0.0010, 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(0.0026, 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(0.0015, 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(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-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(0.0009, 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(0.0015, 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(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(0.0029, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4674e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0007, 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(0.0009, 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(0.0019, 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(0.0037, 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(0.0020, 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(0.0019, 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(0.0009, 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(0.0011, 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(0.0062, 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(0.0017, 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(0.0014, 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(0.0007, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, 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(0.0001, 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(0.0002, 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(0.0023, 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(0.0007, 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(0.0012, 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(0.0003, 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(0.0001, 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(0.0019, 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(0.0012, 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(0.0004, 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(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4572e-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(0.0001, 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(0.0001, 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(0.0005, 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(0.0003, 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(0.0002, 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(0.0019, 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(0.0038, 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(0.0004, 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(0.0008, 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(0.0002, 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(0.0022, 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(0.0023, 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(0.0021, 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(0.0014, 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(0.0002, 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(0.0029, 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(0.0037, 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(0.0016, 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(0.0001, 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(0.0001, 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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<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(0.0039, 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(0.0010, 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(0.0132, 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(0.0005, 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(0.0035, 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(0.0003, 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(0.0006, 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(0.0006, 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(0.0025, 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(0.0001, 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(0.0022, 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(0.0098, device='cuda:0', grad_fn=<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(0.0008, 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(0.0003, 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(0.0005, 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(0.0008, 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(0.0038, 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(0.0020, 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(0.0038, 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(0.0013, 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(0.0007, 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(0.0013, 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(0.0019, 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(0.0008, 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(0.0014, 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(0.0022, 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(0.0001, 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(0.0030, 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(0.0009, 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(0.0017, 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(0.0014, 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(0.0033, device='cuda:0', grad_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(0.0031, 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(0.0021, 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(0.0005, 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(0.0001, 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(0.0034, 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(0.0002, 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(0.0042, 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(0.0007, 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(0.0035, 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(0.0001, 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(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(0.0073, 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(0.0014, 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(0.0011, 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(0.0016, 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(0.0010, 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(0.0003, 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(0.0030, 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(0.0039, 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(0.0009, 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(0.0011, 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(0.0590, 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(0.0045, 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(0.0032, 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(0.0017, 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(0.0022, 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(0.0017, 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(0.0005, 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(9.5037e-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(0.0001, 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(0.0028, 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(0.0022, 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(0.0016, 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(0.0009, 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(0.0021, 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(0.0016, 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(0.0002, 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(0.0016, 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(0.0002, 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(0.0036, 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(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4835e-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(0.0028, 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(0.0015, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, 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(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, 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(0.0017, 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(0.0019, 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(0.0023, 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(0.0006, 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(0.0020, 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(0.0068, 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(0.0008, 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>)
			------------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
				 {'precision': {'support': 0.8899328859060402, 'agreement': 0.4835680751173709, 'direct_attack': 0.20224719101123595, 'undercutter_attack': 0.26104417670682734, 'partial': 0.25}, 'recall': {'support': 0.7700348432055749, 'agreement': 0.6280487804878049, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.15555555555555556}, 'f1': {'support': 0.8256537982565381, 'agreement': 0.5464190981432361, 'direct_attack': 0.17560975609756094, 'undercutter_attack': 0.3513513513513514, 'partial': 0.1917808219178082}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6383136094674556, 'recall': 0.6383136094674556, 'f1': 0.6383136094674556, 'support': None}, 'macro_avg': {'precision': 0.4173584657482949, 'recall': 0.4492003351373334, 'f1': 0.4181629651532989, 'support': None}, 'weighted_avg': {'precision': 0.682754732709452, 'recall': 0.6383136094674556, 'f1': 0.65136477188447, 'support': None}}
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(0.0008, 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(0.0027, 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(0.0052, 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(0.0019, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7155e-05, 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(0.0032, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-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(0.0007, 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(0.0009, 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(0.0010, 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(0.0001, 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(0.0016, 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(0.0011, 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(0.0011, 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(0.0009, 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(9.4311e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9003e-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(0.0023, 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(0.0026, 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(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(0.0023, 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(9.4190e-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(0.0016, 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(0.0005, 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(0.0016, 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(0.0002, 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(0.0022, 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(0.0013, 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(0.0002, 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(0.0002, 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(0.0016, 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(0.0002, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8801e-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(0.0012, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6994e-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(0.0010, 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(0.0008, 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(0.0004, 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(0.0048, 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(0.0003, 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(0.0010, 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(0.0017, 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(7.5238e-05, 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(0.0001, 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(0.0027, 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(0.0018, 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(0.0009, 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(0.0024, 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(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(0.0004, 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(9.4977e-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(0.0021, 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(0.0040, 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(0.0015, 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(0.0026, 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(0.0167, 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(0.0065, 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(0.0017, 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(0.0015, 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(0.0015, 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(0.0009, 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(0.0008, 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(0.0012, 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(8.4653e-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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9790e-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(0.0022, 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(0.0016, 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(0.0001, 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(0.0024, 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(0.0002, 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(0.0014, 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(9.4764e-05, device='cuda:0', grad_fn=<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(0.0004, 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(9.3795e-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(0.0012, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2262e-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(0.0019, 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(0.0023, 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(0.0003, 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(0.0004, 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(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(0.0015, 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(0.0020, 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(9.1374e-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(0.0040, 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(0.0014, 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(0.0054, 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(0.0008, 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(0.0005, 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(0.0015, 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(0.0003, 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(0.0007, 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(0.0015, 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(0.0009, 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(0.0035, 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(0.0021, 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(0.0002, 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(0.0014, 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(0.0006, 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(0.0014, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2988e-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(0.0014, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9073e-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(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(0.0002, 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(0.0009, 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(0.0008, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, 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(0.0044, device='cuda:0', grad_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(0.0008, 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(0.0024, 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(0.0007, 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(0.0013, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5329e-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(9.1707e-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(0.0001, 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(0.0016, 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(0.0020, 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(0.0025, 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(0.0014, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2706e-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(0.0016, 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(0.0018, 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(0.0012, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9588e-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(6.9486e-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(0.0024, 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(0.0096, 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(0.0008, 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(0.0004, 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(0.0001, 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(0.0008, 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(0.0025, 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(0.0002, 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(0.0081, 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(0.0037, 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(0.0008, 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(0.0012, 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(0.0030, device='cuda:0', grad_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(0.0026, 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(0.0002, 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(0.0002, 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(0.0018, 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(9.4190e-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(0.0018, 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(0.0016, 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(0.0015, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-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(0.0014, 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(0.0017, 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(0.0065, 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(0.0035, 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(0.0003, 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(0.0003, 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(0.0012, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3897e-05, device='cuda:0', grad_fn=<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(7.9840e-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(0.0029, 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(0.0007, 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(0.0002, 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(0.0013, 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(0.0015, 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(8.6409e-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(0.0008, 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(0.0027, 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(0.0016, 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(0.0047, 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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5037e-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(9.3402e-05, device='cuda:0', grad_fn=<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(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-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(0.0007, 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(0.0032, 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(0.0015, 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(0.0012, 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(0.0005, 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(0.0003, 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(8.7772e-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(0.0073, 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(0.0006, 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(9.5734e-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(0.0005, 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(0.0004, 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(0.0006, 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(0.0057, 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(0.0010, 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(7.6328e-05, 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(0.0005, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5884e-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(6.8517e-05, device='cuda:0', grad_fn=<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(0.0015, 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(0.0005, 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(0.0054, 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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, 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(0.0010, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3927e-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(0.0008, 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(0.0003, 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(9.6248e-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(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(0.0006, 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(0.0018, 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(0.0003, 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(0.0016, 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(8.7922e-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(0.0071, 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(0.0027, 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(0.0015, 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(0.0008, 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(0.0001, 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(0.0001, 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(0.0005, 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(0.0035, 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(0.0016, 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(0.0017, 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(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(0.0016, 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(0.0015, 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(0.0026, 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(0.0026, 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(0.0007, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4038e-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(0.0004, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2230e-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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6913e-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(0.0022, 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(0.0036, 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(0.0001, 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(0.0010, 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(0.0009, 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(0.0029, 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(0.0016, 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(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(0.0006, 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(0.0006, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8145e-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(0.0014, 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(0.0021, 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(0.0023, 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(8.8710e-05, device='cuda:0', grad_fn=<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(9.9730e-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(0.0011, 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(0.0005, 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(0.0016, 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(0.0030, 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(0.0018, 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(0.0002, 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(0.0005, 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(0.0003, 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(0.0002, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5500e-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(0.0002, 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(0.0008, 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
				 {'precision': {'support': 0.8900804289544236, 'agreement': 0.4855769230769231, 'direct_attack': 0.20224719101123595, 'undercutter_attack': 0.26, 'partial': 0.23728813559322035}, 'recall': {'support': 0.7711962833914053, 'agreement': 0.6158536585365854, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.15555555555555556}, 'f1': {'support': 0.8263845675171126, 'agreement': 0.5430107526881721, 'direct_attack': 0.17560975609756094, 'undercutter_attack': 0.3504043126684636, 'partial': 0.1879194630872483}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6375739644970414, 'recall': 0.6375739644970414, 'f1': 0.6375739644970414, 'support': None}, 'macro_avg': {'precision': 0.4150385357271606, 'recall': 0.4469935987842555, 'f1': 0.41666577041171154, 'support': None}, 'weighted_avg': {'precision': 0.6821527152922096, 'recall': 0.6375739644970414, 'f1': 0.6510749121975945, 'support': None}}
Loss: tensor(0.0007, 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(0.0001, 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(0.0024, 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(0.0037, 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(0.0009, 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(0.0018, 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(0.0023, 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(0.0005, 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(0.0010, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-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(0.0009, 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(0.0011, 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(0.0005, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, 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(0.0023, 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(0.0085, 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(0.0013, 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(0.0035, 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(0.0020, 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(0.0009, 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(0.0017, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3524e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0006, 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(0.0004, 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(0.0009, 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(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(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(0.0015, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5613e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2131e-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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7741e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4795e-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(0.0007, 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(0.0001, 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(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(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8800e-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(8.1142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7005e-05, 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(0.0004, 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(8.9134e-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(7.5268e-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(0.0003, 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(9.3554e-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(0.0008, 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(0.0007, 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(0.0015, 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(0.0009, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5259e-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(0.0012, 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(0.0030, 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(0.0023, 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(0.0012, 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(0.0007, 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(8.5652e-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(0.0003, 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(0.0006, 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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9899e-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(0.0002, 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(0.0012, 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(0.0008, 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(9.8791e-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(0.0021, 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(0.0007, 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(0.0004, 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(9.8821e-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(0.0011, 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(0.0018, 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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9617e-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(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1980e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9839e-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(9.7035e-05, device='cuda:0', grad_fn=<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(0.0020, 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(0.0037, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4330e-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(0.0011, 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(0.0029, device='cuda:0', grad_fn=<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(0.0006, 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(0.0010, 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(0.0011, 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(0.0005, 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(8.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2504e-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(0.0003, 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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7277e-05, 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(0.0019, 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(8.3685e-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(0.0014, 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(0.0020, 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(0.0003, 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(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(0.0024, 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(0.0006, 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(0.0006, 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(0.0005, 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(0.0044, 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(0.0005, 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(0.0013, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-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(0.0004, 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(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(0.0032, 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(0.0012, 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(0.0007, 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(0.0011, 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(0.0006, 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(0.0014, 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(0.0023, 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(0.0005, 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(0.0004, 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(0.0033, 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(0.0003, 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(0.0008, 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(0.0005, 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(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5238e-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(9.8579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-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(0.0018, 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(0.0017, 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(0.0020, device='cuda:0', grad_fn=<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(0.0017, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7811e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-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(0.0003, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9164e-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(0.0006, 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(5.5620e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8286e-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(0.0019, 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(0.0011, 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(0.0009, 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(0.0007, 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(0.0008, 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(0.0003, 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(0.0014, 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(0.0024, 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(0.0009, 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(6.1735e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, 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(0.0011, 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(0.0004, 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(0.0017, 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(0.0002, 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(0.0017, 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(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(0.0013, 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(0.0026, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1929e-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(0.0008, 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(0.0031, 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(0.0003, 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(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4823e-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(9.1677e-05, 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(0.0001, 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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1595e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3382e-05, device='cuda:0', grad_fn=<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(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(0.0001, 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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7811e-05, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, 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(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4542e-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(0.0049, 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(0.0008, 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(0.0019, 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(0.0012, 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(9.3070e-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(0.0019, 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(0.0015, 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(0.0001, 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(0.0004, 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(0.0019, 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(0.0001, 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(0.0012, 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(0.0008, 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(0.0001, 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(8.0173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0345e-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(0.0025, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8356e-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(0.0018, 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(0.0002, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-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(0.0011, 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(0.0008, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-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(0.0016, 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(0.0001, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-05, device='cuda:0', grad_fn=<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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-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(0.0005, 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(0.0011, 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(0.0016, 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(0.0002, 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(0.0020, 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(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(0.0025, 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(0.0008, 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(0.0007, 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(0.0016, 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(6.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6742e-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(0.0037, 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(0.0006, 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(0.0010, 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(0.0019, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6551e-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(9.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-05, device='cuda:0', grad_fn=<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(0.0018, 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(0.0015, 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(0.0007, 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(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-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(0.0019, 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(6.2099e-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(0.0010, 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(0.0001, 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(0.0004, 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(8.2231e-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(0.0025, 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(0.0004, 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(0.0005, 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(0.0001, 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(0.0006, 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(6.0585e-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(0.0024, 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(0.0001, 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(0.0014, 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(0.0003, 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(0.0001, 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(7.8931e-05, device='cuda:0', grad_fn=<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(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(0.0006, 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(6.2250e-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(0.0003, 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>)
			------------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
				 {'precision': {'support': 0.8903743315508021, 'agreement': 0.49029126213592233, 'direct_attack': 0.1978021978021978, 'undercutter_attack': 0.26104417670682734, 'partial': 0.2413793103448276}, 'recall': {'support': 0.7735191637630662, 'agreement': 0.6158536585365854, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.15555555555555556}, 'f1': {'support': 0.8278433809819763, 'agreement': 0.545945945945946, 'direct_attack': 0.17391304347826086, 'undercutter_attack': 0.3513513513513514, 'partial': 0.18918918918918917}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6390532544378699, 'recall': 0.6390532544378699, 'f1': 0.6390532544378699, 'support': None}, 'macro_avg': {'precision': 0.4161782557081154, 'recall': 0.44745817485858774, 'f1': 0.4176485821893447, 'support': None}, 'weighted_avg': {'precision': 0.6828961573322098, 'recall': 0.6390532544378699, 'f1': 0.6523836832430737, 'support': None}}
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(0.0012, 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(9.1132e-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(7.5480e-05, device='cuda:0', grad_fn=<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(7.8508e-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(0.0019, 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(0.0002, 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(0.0008, 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(0.0006, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6831e-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(0.0019, 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(0.0029, 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(0.0013, 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(0.0024, 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(7.8962e-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(0.0016, 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(6.6912e-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(0.0030, 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(0.0004, 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(0.0001, 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(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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-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(0.0004, 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(0.0006, 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(0.0006, 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(0.0015, 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(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(0.0010, 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(0.0025, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4542e-05, 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(0.0012, 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(0.0005, 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(7.6812e-05, 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(0.0002, 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(0.0017, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4068e-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(9.0072e-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(0.0010, 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(7.9234e-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(0.0001, 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(0.0012, 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(0.0005, 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(0.0003, 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(9.6974e-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(0.0011, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1787e-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(0.0010, 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(0.0002, 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(8.3381e-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(0.0017, 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(5.4772e-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(0.0003, 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(0.0010, 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(0.0009, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3543e-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(0.0014, 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(0.0005, 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(0.0025, 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(0.0003, 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(0.0007, 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(0.0007, 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(0.0002, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9931e-05, device='cuda:0', grad_fn=<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(0.0028, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7438e-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(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3806e-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(0.0011, 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(0.0009, 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(6.7427e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, 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(6.2856e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-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(0.0008, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7327e-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(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(0.0019, 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(0.0010, 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(0.0007, 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(0.0003, 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(7.3210e-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(0.0003, 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(0.0012, 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(7.7266e-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(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3037e-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(0.0007, 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(0.0005, 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(9.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2736e-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(0.0025, 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(0.0014, 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(8.5350e-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(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-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(0.0003, 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(8.8559e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8589e-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(8.5318e-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(0.0023, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7336e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-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(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-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(0.0003, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8428e-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(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2837e-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(0.0010, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8477e-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(0.0025, 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(9.9124e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-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(8.6651e-05, 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(0.0001, 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(8.6621e-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(0.0008, 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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0018, 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(0.0031, 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(0.0004, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3986e-05, device='cuda:0', grad_fn=<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(0.0018, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5733e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0006, 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(0.0012, 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(0.0003, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-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(0.0007, 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(0.0010, 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(0.0003, 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(0.0007, 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(0.0017, 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(0.0017, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0020, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8275e-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(0.0032, 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(7.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9013e-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(0.0013, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7438e-05, device='cuda:0', grad_fn=<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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9809e-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(0.0004, 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(0.0001, 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(0.0006, 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(0.0007, 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(5.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(8.3563e-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(6.2856e-05, 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(0.0012, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3098e-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(0.0004, 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(0.0019, 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(0.0010, 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(9.7881e-05, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-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(0.0096, 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(0.0012, 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(0.0012, 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(0.0011, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6266e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4229e-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(0.0004, 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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0009, 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(0.0008, 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(0.0009, 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(0.0022, 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(6.2250e-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(9.4552e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-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(0.0018, 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(0.0004, 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(0.0005, 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(0.0013, 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(0.0011, 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(0.0017, 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(0.0012, 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(0.0006, 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(8.9225e-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(7.4965e-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(0.0006, 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(0.0002, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8113e-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(6.5520e-05, 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(0.0032, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1029e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0010, 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(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8306e-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(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(0.0003, 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(8.6349e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6822e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-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(0.0002, 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(0.0002, 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(0.0009, 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(0.0012, 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(0.0003, 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(7.5965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2331e-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(8.2019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-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(7.5692e-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(0.0010, 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(0.0015, 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(8.1747e-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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-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(0.0007, 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(0.0009, 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(0.0013, 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(0.0001, 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(0.0023, 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(4.9595e-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(0.0006, 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(0.0005, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(7.0001e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0224e-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(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(0.0003, 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(6.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8044e-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(0.0007, 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>)
			------------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
				 {'precision': {'support': 0.890520694259012, 'agreement': 0.48792270531400966, 'direct_attack': 0.1978021978021978, 'undercutter_attack': 0.2620967741935484, 'partial': 0.24561403508771928}, 'recall': {'support': 0.7746806039488966, 'agreement': 0.6158536585365854, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.15555555555555556}, 'f1': {'support': 0.8285714285714285, 'agreement': 0.5444743935309972, 'direct_attack': 0.17391304347826086, 'undercutter_attack': 0.3523035230352304, 'partial': 0.19047619047619047}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.639792899408284, 'recall': 0.639792899408284, 'f1': 0.639792899408284, 'support': None}, 'macro_avg': {'precision': 0.4167912813312974, 'recall': 0.44769046289575376, 'f1': 0.41794771581842144, 'support': None}, 'weighted_avg': {'precision': 0.6830781576988727, 'recall': 0.639792899408284, 'f1': 0.6528397167253563, 'support': None}}
Loss: tensor(7.8901e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1859e-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(0.0004, 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(6.4127e-05, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5228e-05, device='cuda:0', grad_fn=<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(8.2806e-05, device='cuda:0', grad_fn=<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(6.8790e-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(0.0002, 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(0.0020, 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(0.0005, 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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0016, 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(0.0005, 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(0.0013, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2775e-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(7.8356e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6127e-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(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9062e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9346e-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(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9143e-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(0.0016, 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(0.0008, 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(0.0015, 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(0.0005, 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(0.0016, 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(9.2373e-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(8.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-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(0.0019, 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(8.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-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(6.8970e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5470e-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(7.7448e-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(9.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-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(0.0024, 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(0.0016, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-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(4.8232e-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(5.3803e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2221e-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(0.0019, 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(0.0006, 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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-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(0.0015, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6076e-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(0.0010, 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(5.9041e-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(5.2683e-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(0.0004, 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(0.0001, 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(0.0002, 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(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3643e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<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(8.3079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5420e-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(0.0006, 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(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0041e-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(0.0012, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-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(0.0004, 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(0.0019, 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(0.0002, 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(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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5208e-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(9.7004e-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(0.0011, 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(0.0014, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-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(0.0003, 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(0.0012, 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(0.0015, 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(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7933e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0361e-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(0.0012, 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(0.0004, 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(0.0001, 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(8.0869e-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(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5227e-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(5.3137e-05, device='cuda:0', grad_fn=<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(0.0054, device='cuda:0', grad_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(0.0010, 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(0.0021, 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(0.0010, 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(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(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(0.0003, 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(0.0006, 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(0.0014, 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(7.7447e-05, 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(0.0001, 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(0.0007, 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(0.0002, 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(7.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, 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(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1596e-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(0.0011, 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(8.7348e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1717e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, 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(0.0009, 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(0.0003, 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(0.0001, 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(0.0015, 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(6.2825e-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(0.0007, 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(0.0001, 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(0.0014, 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(0.0003, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0171e-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(0.0006, 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(0.0008, 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(0.0010, 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(0.0017, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-05, 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(0.0012, 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(0.0010, 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(6.1644e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5056e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-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(0.0009, 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(0.0002, 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(7.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0002, 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(0.0009, 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(0.0009, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, 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(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-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(0.0006, 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(0.0030, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<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(0.0008, 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(8.6861e-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(0.0014, 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(0.0009, 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(7.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-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(0.0008, 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(0.0002, 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(0.0004, 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(0.0012, 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(0.0005, 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(0.0013, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8850e-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(8.1868e-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(0.0003, 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(0.0010, 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(0.0014, 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(0.0016, 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(9.5703e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9799e-05, device='cuda:0', grad_fn=<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(4.6900e-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(0.0004, 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(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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-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(7.1211e-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(0.0003, 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(6.5066e-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(0.0005, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6216e-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(0.0003, 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(0.0002, 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(6.2765e-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(6.3915e-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(7.7024e-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(9.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9365e-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(0.0007, 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(0.0001, 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(0.0003, 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(0.0007, 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(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(6.8759e-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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2603e-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(0.0006, 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(0.0015, 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(5.5589e-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(9.1160e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-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(8.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-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(0.0008, 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(0.0003, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0999e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0046, 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(0.0002, 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(0.0016, 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(0.0005, 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(0.0022, device='cuda:0', grad_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(0.0010, 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(0.0010, 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(0.0002, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8608e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1495e-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(0.0009, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9224e-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(0.0014, 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(9.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6137e-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(0.0005, 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(5.8072e-05, 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(0.0002, 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(0.0003, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, 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(0.0010, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1716e-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(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-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(9.9518e-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(9.1223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9345e-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(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-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(0.0007, 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(7.9658e-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(0.0025, 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(9.4522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8063e-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(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(0.0003, 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(0.0002, 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(0.0021, device='cuda:0', grad_fn=<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(0.0001, 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(8.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-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
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.891566265060241, 'agreement': 0.48792270531400966, 'direct_attack': 0.1935483870967742, 'undercutter_attack': 0.26104417670682734, 'partial': 0.25}, 'recall': {'support': 0.7735191637630662, 'agreement': 0.6158536585365854, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.15555555555555556}, 'f1': {'support': 0.8283582089552239, 'agreement': 0.5444743935309972, 'direct_attack': 0.17224880382775118, 'undercutter_attack': 0.3513513513513514, 'partial': 0.1917808219178082}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6390532544378699, 'recall': 0.6390532544378699, 'f1': 0.6390532544378699, 'support': None}, 'macro_avg': {'precision': 0.4168163068355704, 'recall': 0.44745817485858774, 'f1': 0.4176427159166264, 'support': None}, 'weighted_avg': {'precision': 0.683576801903193, 'recall': 0.6390532544378699, 'f1': 0.6525627715825938, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0005, 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(0.0007, 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(7.1271e-05, device='cuda:0', grad_fn=<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(7.0485e-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(8.9164e-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(7.1787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-05, 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(5.3712e-05, device='cuda:0', grad_fn=<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(0.0006, 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(0.0010, 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(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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6287e-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(6.8911e-05, 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(0.0014, device='cuda:0', grad_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(0.0006, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7862e-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(0.0009, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9737e-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(5.9374e-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(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(7.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0054, 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(9.3826e-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(0.0004, 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(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(0.0010, 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(6.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9143e-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(0.0007, 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(0.0011, 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(0.0014, 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(0.0007, 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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, device='cuda:0', grad_fn=<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(0.0014, 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(0.0011, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5764e-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(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0253e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6399e-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(0.0018, 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(0.0008, 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(6.5338e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0434e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0011, 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(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5228e-05, device='cuda:0', grad_fn=<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(7.9143e-05, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6185e-05, device='cuda:0', grad_fn=<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(0.0004, 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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-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(0.0001, 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(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-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(6.0706e-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(0.0067, 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(5.2562e-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(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-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(0.0010, 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(0.0002, 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(6.6307e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-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(5.1956e-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(0.0009, 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(8.0052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2806e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6479e-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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4288e-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(8.1595e-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(0.0006, 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(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-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(7.4360e-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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0008, 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(7.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7044e-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(0.0006, 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(0.0010, 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(0.0006, 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(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5592e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4309e-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(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-05, device='cuda:0', grad_fn=<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(0.0013, 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(0.0006, 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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8850e-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(0.0003, 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(0.0009, 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(0.0008, 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(5.4409e-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(0.0007, 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(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4290e-05, device='cuda:0', grad_fn=<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(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-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(0.0009, 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(0.0003, 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(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(0.0030, 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(0.0008, 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(0.0013, 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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5014e-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(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-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(0.0013, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4300e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-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(0.0008, 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(0.0052, device='cuda:0', grad_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(0.0004, 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(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1533e-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(7.4663e-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(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5016e-05, device='cuda:0', grad_fn=<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(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-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(6.4914e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5299e-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(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-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(0.0001, 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(0.0006, 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(8.6712e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1826e-05, 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(0.0005, 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(0.0015, 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(0.0004, 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(0.0005, 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(5.1199e-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(0.0020, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-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(0.0003, 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(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(4.9837e-05, device='cuda:0', grad_fn=<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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-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(6.7215e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0014, 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(0.0007, 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(6.7336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, device='cuda:0', grad_fn=<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(0.0010, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-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(8.2897e-05, 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(0.0005, 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(0.0008, 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(0.0011, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7770e-05, device='cuda:0', grad_fn=<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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0005, 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(0.0004, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-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(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(0.0005, 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(5.2229e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-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(0.0001, device='cuda:0', grad_fn=<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(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<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(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-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(0.0013, 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(0.0006, 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(4.9262e-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(7.7145e-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(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(7.4875e-05, 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(0.0004, 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(7.0424e-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(0.0006, 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(0.0015, 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(5.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8820e-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(6.5974e-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(6.5701e-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(7.0091e-05, device='cuda:0', grad_fn=<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(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(6.6004e-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(4.6900e-05, 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(0.0022, 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(0.0012, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7962e-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(0.0009, 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(0.0001, 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(0.0002, 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(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, 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(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(9.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-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(0.0013, 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(6.7094e-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(0.0013, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4189e-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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-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(0.0003, 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(0.0003, 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(0.0016, 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(0.0009, 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(5.3258e-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(0.0011, 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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2945e-05, 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(4.7839e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3956e-05, device='cuda:0', grad_fn=<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(0.0005, 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>)
			------------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
				 {'precision': {'support': 0.891566265060241, 'agreement': 0.49038461538461536, 'direct_attack': 0.1935483870967742, 'undercutter_attack': 0.2631578947368421, 'partial': 0.2631578947368421}, 'recall': {'support': 0.7735191637630662, 'agreement': 0.6219512195121951, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.16666666666666666}, 'f1': {'support': 0.8283582089552239, 'agreement': 0.5483870967741936, 'direct_attack': 0.17224880382775118, 'undercutter_attack': 0.3532608695652174, 'partial': 0.2040816326530612}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6405325443786982, 'recall': 0.6405325443786982, 'f1': 0.6405325443786982, 'support': None}, 'macro_avg': {'precision': 0.420363011403063, 'recall': 0.4508999092759319, 'f1': 0.4212673223550894, 'support': None}, 'weighted_avg': {'precision': 0.6849405028347958, 'recall': 0.6405325443786982, 'f1': 0.6540271266136106, 'support': None}}
Loss: tensor(0.0008, 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(5.6861e-05, 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(5.3743e-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(0.0004, 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(4.6325e-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(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(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(4.9716e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6589e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8759e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4733e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-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(6.2250e-05, 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(0.0002, 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(8.3048e-05, device='cuda:0', grad_fn=<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(5.8557e-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(6.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6912e-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(0.0005, 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(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(0.0006, 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(7.4874e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-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(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(0.0008, 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(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-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(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-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(0.0012, 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(6.8547e-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(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(0.0006, 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(0.0006, 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(4.6446e-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(8.2655e-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(5.2380e-05, device='cuda:0', grad_fn=<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(6.8305e-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(0.0013, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-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(0.0004, 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(0.0002, 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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4067e-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(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1847e-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(0.0007, 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(0.0003, 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(8.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9204e-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(0.0004, 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(7.5359e-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(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6893e-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(4.3237e-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(6.4157e-05, device='cuda:0', grad_fn=<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(7.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-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(0.0003, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-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(0.0008, 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(8.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5017e-05, 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(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(0.0010, 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(0.0002, 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(8.3503e-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(0.0005, 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(0.0003, 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(6.8062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-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(0.0001, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-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(6.6367e-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(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-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(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(0.0008, 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(5.9737e-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(0.0007, 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(0.0008, 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(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8758e-05, 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(0.0005, 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(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-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(0.0006, 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(6.0464e-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(0.0010, 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(0.0002, 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(4.6295e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4856e-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(0.0020, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-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(7.4148e-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(0.0001, 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(5.8042e-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(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-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(5.8708e-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(5.6679e-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(8.7831e-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(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(0.0002, device='cuda:0', grad_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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1283e-05, 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(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(7.1211e-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(0.0008, 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(0.0012, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-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(0.0005, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1454e-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(9.4974e-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(0.0003, 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(0.0004, 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(0.0017, 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(5.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-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(9.6853e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9565e-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(6.7518e-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(4.5114e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-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(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(0.0004, 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(0.0055, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-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(0.0001, 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(0.0012, 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(0.0006, 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(6.5731e-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(0.0009, 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(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5479e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6873e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5429e-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(4.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-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(4.4599e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1180e-05, device='cuda:0', grad_fn=<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(3.9634e-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(0.0005, 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(0.0010, 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(7.8295e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-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(7.3876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-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(0.0007, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5640e-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(0.0009, 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(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1524e-05, 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(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(0.0002, 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(9.2945e-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(5.2532e-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(0.0016, 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(9.8850e-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(0.0011, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-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(0.0006, 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(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7255e-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(0.0008, 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(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1484e-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(6.1160e-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(5.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-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(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0919e-05, device='cuda:0', grad_fn=<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(9.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, 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(0.0002, 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(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(4.1632e-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(0.0010, 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(0.0001, 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(0.0003, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0424e-05, 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(5.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2180e-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(0.0005, 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(0.0004, 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(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-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(0.0005, 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(5.1411e-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(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7488e-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(0.0003, 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(0.0003, 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(0.0006, 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(0.0001, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-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(8.5229e-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(4.3782e-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(0.0009, 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(8.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-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(0.0008, 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(0.0004, 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(8.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-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(0.0010, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1707e-05, device='cuda:0', grad_fn=<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(0.0017, 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(0.0007, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<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(0.0016, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6034e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2160e-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(0.0002, 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(4.2086e-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(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(0.0004, 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(0.0003, 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(8.0051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5429e-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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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
				 {'precision': {'support': 0.891566265060241, 'agreement': 0.49282296650717705, 'direct_attack': 0.19148936170212766, 'undercutter_attack': 0.26422764227642276, 'partial': 0.26785714285714285}, 'recall': {'support': 0.7735191637630662, 'agreement': 0.6280487804878049, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.16666666666666666}, 'f1': {'support': 0.8283582089552239, 'agreement': 0.552278820375335, 'direct_attack': 0.1714285714285714, 'undercutter_attack': 0.3542234332425068, 'partial': 0.20547945205479448}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6412721893491125, 'recall': 0.6412721893491125, 'f1': 0.6412721893491125, 'support': None}, 'macro_avg': {'precision': 0.42159267568062225, 'recall': 0.4521194214710538, 'f1': 0.42235369721128635, 'support': None}, 'weighted_avg': {'precision': 0.6854681762234329, 'recall': 0.6412721893491125, 'f1': 0.6546080213350531, 'support': None}}
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(6.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7517e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7730e-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(6.4823e-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(0.0005, 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(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(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4713e-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(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-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(0.0009, 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(6.8244e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<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(7.6569e-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(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5710e-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(0.0005, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(7.6509e-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(6.3976e-05, device='cuda:0', grad_fn=<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(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, 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(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(0.0003, 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(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0004, 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(3.8907e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6913e-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(0.0010, 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(8.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4601e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-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(5.0261e-05, device='cuda:0', grad_fn=<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(3.9149e-05, device='cuda:0', grad_fn=<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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0008, 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(0.0003, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4632e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-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(0.0022, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9881e-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(9.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(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(6.4852e-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(0.0001, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4500e-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(4.1844e-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(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(0.0004, 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(6.4460e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1948e-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(0.0003, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7336e-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(4.1511e-05, device='cuda:0', grad_fn=<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(0.0004, 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(9.0676e-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(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-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(0.0005, 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(0.0010, 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(0.0006, 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(5.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-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(0.0002, 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(0.0003, 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(8.1232e-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(0.0008, 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(5.1260e-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(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4369e-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(7.4572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, 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(5.6498e-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(0.0010, 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(0.0003, 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(0.0003, 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(0.0007, 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(0.0002, 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(9.1919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8748e-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(0.0006, 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(0.0011, 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(4.3509e-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(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(4.8777e-05, device='cuda:0', grad_fn=<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(4.4993e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1514e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-05, 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(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0009, 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(0.0002, 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(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-05, 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(0.0001, 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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-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(6.7609e-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(0.0007, 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(7.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1715e-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(0.0017, 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(0.0039, 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(0.0011, 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(0.0001, 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(0.0005, 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(4.7808e-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(3.3942e-05, 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(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8759e-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(4.0663e-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(0.0003, 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(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-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(0.0001, 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(3.2791e-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(0.0002, 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(0.0004, 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(3.6546e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0133e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-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(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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9586e-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(0.0011, 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(0.0009, 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(5.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4925e-05, 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(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(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0404e-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(4.0784e-05, device='cuda:0', grad_fn=<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(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, 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(4.9806e-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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, 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(0.0004, 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(0.0007, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3543e-05, 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(5.6407e-05, device='cuda:0', grad_fn=<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(7.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-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(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-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(4.5235e-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(0.0002, 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(0.0009, 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(3.1217e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8041e-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(0.0023, 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(7.1696e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5520e-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(4.2995e-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(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(0.0001, 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(0.0024, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9618e-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(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-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(7.0575e-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(7.4632e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-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(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(3.9301e-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(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-05, device='cuda:0', grad_fn=<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(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0001, 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(6.3491e-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(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7448e-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(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(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-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(5.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-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(6.4823e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-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(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-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(8.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-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(4.6840e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, 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(0.0004, 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(0.0005, 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(0.0004, 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(0.0002, 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(5.4621e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5045e-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(5.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4663e-05, device='cuda:0', grad_fn=<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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1613e-05, 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(0.0006, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8204e-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>)
			------------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
				 {'precision': {'support': 0.8918558077436582, 'agreement': 0.4975845410628019, 'direct_attack': 0.1935483870967742, 'undercutter_attack': 0.26422764227642276, 'partial': 0.2807017543859649}, 'recall': {'support': 0.775842044134727, 'agreement': 0.6280487804878049, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8298136645962733, 'agreement': 0.555256064690027, 'direct_attack': 0.17224880382775118, 'undercutter_attack': 0.3542234332425068, 'partial': 0.21768707482993196}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.643491124260355, 'recall': 0.643491124260355, 'f1': 0.643491124260355, 'support': None}, 'macro_avg': {'precision': 0.42558362651312437, 'recall': 0.45480621976760827, 'f1': 0.42584580823729806, 'support': None}, 'weighted_avg': {'precision': 0.6872618570377212, 'recall': 0.643491124260355, 'f1': 0.6567790630381746, 'support': None}}
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(0.0002, 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(9.6032e-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(4.3358e-05, device='cuda:0', grad_fn=<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(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6923e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-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(4.7718e-05, device='cuda:0', grad_fn=<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(0.0012, 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(0.0008, 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(0.0009, 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(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4853e-05, 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(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-05, device='cuda:0', grad_fn=<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(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-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(0.0004, 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(5.1956e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-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(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, 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(0.0005, 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(0.0005, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-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(0.0007, 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(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<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(4.8717e-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(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, 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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-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(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, 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(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-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(4.0875e-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(0.0002, 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(0.0007, 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(5.9162e-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(4.4236e-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(7.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-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(0.0032, 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(0.0004, 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(0.0004, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7075e-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(6.1189e-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(0.0012, 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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4621e-05, 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(0.0008, 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(0.0007, 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(0.0011, 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(0.0008, 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(4.1784e-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(3.4396e-05, device='cuda:0', grad_fn=<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(8.8072e-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(0.0007, 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(7.0787e-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(0.0001, 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(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-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(7.1605e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-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(2.4707e-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(0.0007, 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(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(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-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(5.0897e-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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-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(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-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(4.9746e-05, 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(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3732e-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(3.6667e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-05, 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(0.0004, 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(0.0004, 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(4.5114e-05, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6629e-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(7.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4238e-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(8.9466e-05, 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(0.0005, 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(0.0010, 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(0.0003, 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(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<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(5.1532e-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(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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-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(5.5014e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-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(8.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-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(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-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(0.0002, 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(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-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(0.0006, 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(0.0008, 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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3421e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9816e-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(3.1217e-05, device='cuda:0', grad_fn=<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(0.0013, 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(2.9340e-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(5.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7640e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-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(4.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0475e-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(0.0006, 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(6.8305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0788e-05, 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(0.0003, 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(4.4599e-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(0.0011, 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(8.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9011e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-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(0.0006, 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(8.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0110e-05, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-05, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1422e-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(4.5417e-05, device='cuda:0', grad_fn=<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(3.8907e-05, 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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-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(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7396e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(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(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-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(0.0008, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<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(5.8677e-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(0.0010, 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(5.2138e-05, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4227e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, 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(0.0001, 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(0.0007, 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(0.0005, 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(0.0002, 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(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5459e-05, device='cuda:0', grad_fn=<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(3.2549e-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(7.9324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3086e-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(0.0006, 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(9.7217e-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(0.0011, 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(0.0002, 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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3067e-05, device='cuda:0', grad_fn=<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(0.0011, 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(8.0627e-05, 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(0.0005, 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(8.1897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5509e-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(8.0837e-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(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-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(0.0001, 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(0.0006, 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(5.4560e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, 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(0.0005, 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(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(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7447e-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(5.3803e-05, device='cuda:0', grad_fn=<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(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1121e-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(6.4338e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3914e-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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, 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(9.4791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9940e-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(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246e-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(3.5456e-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(0.0003, 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(7.1757e-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(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, 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(0.0002, 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(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<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(3.7424e-05, device='cuda:0', grad_fn=<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(4.1360e-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(8.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-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(3.7878e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5178e-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(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(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(0.0004, 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(9.0016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1432e-05, device='cuda:0', grad_fn=<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(0.0011, 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(0.0007, 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(0.0006, 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(0.0005, 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(5.7285e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0070e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, 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(0.0003, 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
				 {'precision': {'support': 0.892, 'agreement': 0.4975845410628019, 'direct_attack': 0.1956521739130435, 'undercutter_attack': 0.26422764227642276, 'partial': 0.2807017543859649}, 'recall': {'support': 0.7770034843205574, 'agreement': 0.6280487804878049, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8305400372439479, 'agreement': 0.555256064690027, 'direct_attack': 0.17307692307692307, 'undercutter_attack': 0.3542234332425068, 'partial': 0.21768707482993196}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6442307692307693, 'recall': 0.6442307692307693, 'f1': 0.6442307692307693, 'support': None}, 'macro_avg': {'precision': 0.42603322232764657, 'recall': 0.4550385078047743, 'f1': 0.4261567066166673, 'support': None}, 'weighted_avg': {'precision': 0.687534186034317, 'recall': 0.6442307692307693, 'f1': 0.6573126937205354, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-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(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0006, 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(6.4127e-05, device='cuda:0', grad_fn=<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(0.0020, device='cuda:0', grad_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(0.0005, 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(0.0004, 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(4.6355e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0007, 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(0.0002, 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(6.9697e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-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(0.0016, 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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(3.6788e-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(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0553e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0004, 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(4.0421e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-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(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9922e-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(0.0006, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<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(0.0002, 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(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(0.0005, 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(0.0002, 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(0.0005, 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(0.0003, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, 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(0.0005, 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(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1635e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9709e-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(4.4478e-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(0.0002, 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(0.0003, 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(0.0003, 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(0.0007, 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(0.0007, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6448e-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(0.0005, 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(7.5448e-05, 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(0.0006, 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(0.0002, 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(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0978e-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(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9848e-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(5.1684e-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(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-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(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, 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(0.0033, 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(5.0231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-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(0.0001, 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(0.0003, 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(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-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(5.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9858e-05, device='cuda:0', grad_fn=<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(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0594e-05, 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(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3914e-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(3.8392e-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(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, 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(3.5092e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0003, 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(3.8695e-05, 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(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6386e-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(3.5940e-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(0.0007, 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(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7224e-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(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.0884e-05, device='cuda:0', grad_fn=<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(7.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<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(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3380e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4772e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8980e-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(3.3609e-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(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(0.0006, 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(0.0006, 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(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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-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(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(4.9353e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4944e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-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(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(0.0002, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-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(0.0006, 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(0.0007, 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(0.0004, 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(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7944e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7496e-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(3.3730e-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(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3785e-05, 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(0.0006, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7286e-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(0.0004, 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(5.0140e-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(2.6221e-05, 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(7.7870e-05, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-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(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3611e-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(0.0006, 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(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0221e-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(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-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(4.7597e-05, device='cuda:0', grad_fn=<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(0.0004, 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(5.2199e-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(0.0009, 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(9.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, 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(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6170e-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(0.0001, 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(3.6727e-05, 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(5.6195e-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(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0001, 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(6.7367e-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(3.5153e-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(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(5.1048e-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(8.0383e-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(0.0012, 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(3.4759e-05, 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(0.0002, 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(3.6818e-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(4.8717e-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(6.2249e-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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<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(6.1674e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, 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(0.0001, 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(4.5841e-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(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-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(0.0006, 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(0.0005, 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(9.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-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(3.2367e-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(2.7796e-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(3.3609e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8345e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-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(5.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(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(9.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, 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(0.0003, 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(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(4.0875e-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(7.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, 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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-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(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(7.0636e-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(0.0002, 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(3.9634e-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(7.0241e-05, device='cuda:0', grad_fn=<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(4.8444e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9888e-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(4.1269e-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(0.0007, 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(0.0007, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6022e-05, device='cuda:0', grad_fn=<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(3.1853e-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(4.8141e-05, device='cuda:0', grad_fn=<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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1556e-05, device='cuda:0', grad_fn=<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(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6377e-05, 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(0.0007, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-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
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8908122503328895, 'agreement': 0.4975845410628019, 'direct_attack': 0.1956521739130435, 'undercutter_attack': 0.2653061224489796, 'partial': 0.2807017543859649}, 'recall': {'support': 0.7770034843205574, 'agreement': 0.6280487804878049, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8300248138957816, 'agreement': 0.555256064690027, 'direct_attack': 0.17307692307692307, 'undercutter_attack': 0.35519125683060104, 'partial': 0.21768707482993196}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6442307692307693, 'recall': 0.6442307692307693, 'f1': 0.6442307692307693, 'support': None}, 'macro_avg': {'precision': 0.4260113684287359, 'recall': 0.4550385078047743, 'f1': 0.426247226664653, 'support': None}, 'weighted_avg': {'precision': 0.6868743070679687, 'recall': 0.6442307692307693, 'f1': 0.6570711991579528, 'support': None}}
Loss: tensor(0.0003, 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(0.0003, 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(0.0006, 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(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8289e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-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(6.3249e-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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-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(0.0005, 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(0.0002, 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(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(0.0002, 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(8.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6023e-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(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3663e-05, 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(5.5619e-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(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6894e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4674e-05, device='cuda:0', grad_fn=<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(9.0545e-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(3.0581e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, 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(4.4115e-05, device='cuda:0', grad_fn=<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(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-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(9.6157e-05, 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(0.0003, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, 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(3.4759e-05, 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(0.0004, 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(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0003, 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(0.0006, 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(4.1995e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0697e-05, 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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-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(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-05, 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(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(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4974e-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(5.3651e-05, 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(0.0003, 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(3.0278e-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(3.0642e-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(8.6189e-05, 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(0.0003, 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(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7276e-05, device='cuda:0', grad_fn=<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(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-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(0.0001, 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(3.6182e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4940e-05, 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(2.9461e-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(0.0004, 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(4.2026e-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(0.0003, 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(5.4802e-05, device='cuda:0', grad_fn=<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(0.0032, 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(9.2007e-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(3.1580e-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(4.4206e-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(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-05, 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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7222e-05, 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(9.0011e-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(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-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(0.0004, 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(0.0010, 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(3.2398e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3006e-05, 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(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(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(0.0022, 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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-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(9.9846e-05, 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(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8840e-05, 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(0.0002, 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(7.7539e-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(0.0009, 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(0.0003, 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(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6774e-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(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8748e-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(2.6372e-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(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0672e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-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(3.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<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(3.3064e-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(0.0002, 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(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3186e-05, device='cuda:0', grad_fn=<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(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(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(3.0399e-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(0.0001, 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(0.0003, 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(0.0004, 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(4.1239e-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(0.0002, device='cuda:0', grad_fn=<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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, 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(9.8744e-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(5.0019e-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(5.8828e-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(0.0004, 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(5.4378e-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(9.9865e-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(4.3782e-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(7.4087e-05, 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(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(9.4853e-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(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(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-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(6.5972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-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(2.5585e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6790e-05, device='cuda:0', grad_fn=<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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<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(8.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7477e-05, device='cuda:0', grad_fn=<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(0.0006, 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(6.5307e-05, device='cuda:0', grad_fn=<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(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-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(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-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(0.0005, 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(0.0009, 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(3.5213e-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(8.0201e-05, device='cuda:0', grad_fn=<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(7.9347e-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(3.7333e-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(4.3055e-05, 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(4.5083e-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(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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9615e-05, device='cuda:0', grad_fn=<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(0.0004, 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(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8007e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4439e-05, device='cuda:0', grad_fn=<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(8.7708e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-05, device='cuda:0', grad_fn=<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(2.5434e-05, device='cuda:0', grad_fn=<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(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-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(9.1811e-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(7.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4964e-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(4.0118e-05, device='cuda:0', grad_fn=<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(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-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(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<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(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-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(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<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(3.9543e-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(0.0006, 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(0.0001, 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(0.0001, 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(0.0004, 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(8.8904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8415e-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(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-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(0.0002, device='cuda:0', grad_fn=<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(0.0002, 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(0.0004, 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(3.1944e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, 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(0.0005, 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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-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(8.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-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(4.1450e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<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(0.0005, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, 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(0.0001, 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(0.0003, 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(0.0005, 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(6.1008e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9137e-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(0.0005, 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(0.0003, 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(0.0004, 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(0.0001, 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(0.0001, 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(1.8531e-05, 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(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(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-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(2.6494e-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(6.8577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-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(0.0001, 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(0.0006, 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(0.0004, 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(4.1480e-05, device='cuda:0', grad_fn=<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(0.0008, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0604e-05, 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
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8896276595744681, 'agreement': 0.5, 'direct_attack': 0.1956521739130435, 'undercutter_attack': 0.2653061224489796, 'partial': 0.2909090909090909}, 'recall': {'support': 0.7770034843205574, 'agreement': 0.6341463414634146, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5371900826446281, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8295102293862369, 'agreement': 0.5591397849462365, 'direct_attack': 0.17307692307692307, 'undercutter_attack': 0.35519125683060104, 'partial': 0.2206896551724138}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6449704142011834, 'recall': 0.6449704142011834, 'f1': 0.6449704142011834, 'support': None}, 'macro_avg': {'precision': 0.4282990093691164, 'recall': 0.4562580199998963, 'f1': 0.4275215698824823, 'support': None}, 'weighted_avg': {'precision': 0.6870924009361499, 'recall': 0.6449704142011834, 'f1': 0.6574144721536063, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-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(0.0004, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9607e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0002, 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(0.0022, 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(0.0006, 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(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2775e-05, device='cuda:0', grad_fn=<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(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<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(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<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(3.9755e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1773e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7465e-05, device='cuda:0', grad_fn=<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(8.0062e-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(9.6247e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-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(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(5.4439e-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(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<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(5.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-05, 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(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, 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(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(0.0025, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-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(4.6446e-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(0.0006, 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(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9043e-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(0.0007, 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(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-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(0.0004, 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(3.8059e-05, device='cuda:0', grad_fn=<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(9.6397e-05, 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(3.8756e-05, device='cuda:0', grad_fn=<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(0.0007, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-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(0.0013, 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(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(0.0006, 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(4.2722e-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(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(7.4388e-05, device='cuda:0', grad_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(9.5079e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9457e-05, device='cuda:0', grad_fn=<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(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(5.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3607e-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(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(0.0003, 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(6.7487e-05, 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(4.7203e-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(0.0004, 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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-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(4.9352e-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(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(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4631e-05, 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(0.0004, 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(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-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(3.2791e-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(0.0004, 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(3.7212e-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(0.0005, 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(0.0006, 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(0.0004, 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(0.0001, 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(2.7220e-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(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-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(5.4317e-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(8.3957e-05, device='cuda:0', grad_fn=<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(3.1187e-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(7.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7939e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8385e-05, device='cuda:0', grad_fn=<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(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3067e-05, device='cuda:0', grad_fn=<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(8.0941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-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(2.0620e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9401e-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(0.0003, 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(0.0005, 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(7.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5574e-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(0.0001, 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(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-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(0.0001, 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(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-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(0.0007, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1045e-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(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9011e-05, device='cuda:0', grad_fn=<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(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8065e-05, device='cuda:0', grad_fn=<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(5.3833e-05, device='cuda:0', grad_fn=<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(6.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, 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(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(0.0001, 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(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, 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(2.2224e-05, 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(6.9091e-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(3.5789e-05, device='cuda:0', grad_fn=<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(0.0002, 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(5.8263e-05, device='cuda:0', grad_fn=<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(2.7372e-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(2.3284e-05, 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(3.0853e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-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(3.6273e-05, device='cuda:0', grad_fn=<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(6.1850e-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(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9750e-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(0.0003, 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(6.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-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(2.4314e-05, device='cuda:0', grad_fn=<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(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6306e-05, device='cuda:0', grad_fn=<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(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3346e-05, 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(0.0003, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0653e-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(0.0003, 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(2.9764e-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(9.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2897e-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(3.6455e-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(5.0987e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-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(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<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(6.5397e-05, device='cuda:0', grad_fn=<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(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-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(7.7919e-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(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6898e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4780e-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(0.0001, 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(2.8674e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-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(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(8.5468e-05, device='cuda:0', grad_fn=<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(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<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(4.3933e-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(2.6463e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0242e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<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(4.0482e-05, device='cuda:0', grad_fn=<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(2.4737e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5909e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-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(7.5689e-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(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8072e-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(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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5921e-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(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, 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(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(0.0002, 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(6.9166e-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(0.0002, 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(2.6494e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9605e-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(0.0008, 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(8.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<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(0.0002, 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(0.0007, 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(2.5070e-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(6.7546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<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(0.0004, 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(4.6143e-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(9.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, 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(3.8211e-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(7.4542e-05, device='cuda:0', grad_fn=<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(3.1822e-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(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1822e-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(0.0003, 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(0.0004, 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(0.0003, 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(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6969e-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(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9419e-05, device='cuda:0', grad_fn=<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(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-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(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(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-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
				 {'precision': {'support': 0.889920424403183, 'agreement': 0.5048076923076923, 'direct_attack': 0.1956521739130435, 'undercutter_attack': 0.26229508196721313, 'partial': 0.2962962962962963}, 'recall': {'support': 0.7793263646922184, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5289256198347108, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8309597523219814, 'agreement': 0.564516129032258, 'direct_attack': 0.17307692307692307, 'undercutter_attack': 0.3506849315068493, 'partial': 0.2222222222222222}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6464497041420119, 'recall': 0.6464497041420119, 'f1': 0.6464497041420119, 'support': None}, 'macro_avg': {'precision': 0.4297943337774857, 'recall': 0.45628921570736686, 'f1': 0.4282919916320468, 'support': None}, 'weighted_avg': {'precision': 0.6879511617664309, 'recall': 0.6464497041420119, 'f1': 0.6586884553992367, 'support': None}}
Loss: tensor(0.0001, 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(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<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(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7974e-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(4.8081e-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(0.0001, 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(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2402e-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(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(3.4759e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(0.0001, 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(0.0007, 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(0.0003, 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(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(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-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(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8901e-05, device='cuda:0', grad_fn=<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(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-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(6.9952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-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(5.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7192e-05, device='cuda:0', grad_fn=<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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0218e-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(0.0002, 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(2.2891e-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(4.8232e-05, 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(6.7394e-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(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(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(9.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_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(9.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6456e-05, 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(3.1550e-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(0.0002, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7265e-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(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5197e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4632e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4930e-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(0.0004, 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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-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(0.0003, 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(7.1149e-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(2.8401e-05, device='cuda:0', grad_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(3.4699e-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(6.6519e-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(4.0663e-05, device='cuda:0', grad_fn=<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(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6526e-05, 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(0.0001, 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(0.0001, 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(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-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(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(0.0005, 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(0.0001, 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(8.2684e-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(3.3972e-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(0.0002, 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(3.5183e-05, device='cuda:0', grad_fn=<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(5.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6794e-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(0.0003, 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(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<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(0.0002, 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(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8673e-05, device='cuda:0', grad_fn=<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(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4105e-05, 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(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(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-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(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3606e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6742e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-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(0.0006, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3521e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<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(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-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(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(3.5304e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-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(7.2716e-05, 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(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4393e-05, device='cuda:0', grad_fn=<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(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8737e-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(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(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5661e-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(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1062e-05, 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(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, device='cuda:0', grad_fn=<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(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<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(2.4162e-05, 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(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(3.7030e-05, device='cuda:0', grad_fn=<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(2.8522e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-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(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4215e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-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(9.8610e-05, 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(5.4589e-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(8.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9826e-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(3.6243e-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(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-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(3.1550e-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(3.7393e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3172e-05, 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(0.0002, 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(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(9.5731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6510e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6003e-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(2.6494e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8888e-05, 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(8.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, 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(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, 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(2.2951e-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(6.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-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(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-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(5.0079e-05, device='cuda:0', grad_fn=<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(2.8098e-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(3.2670e-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(1.7925e-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(0.0002, 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(4.2662e-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(5.0564e-05, device='cuda:0', grad_fn=<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(9.8152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-05, device='cuda:0', grad_fn=<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(0.0004, 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(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(4.7475e-05, device='cuda:0', grad_fn=<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(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5974e-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(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<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(3.5668e-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(0.0005, 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(2.1074e-05, device='cuda:0', grad_fn=<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(6.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<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(9.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9348e-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(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0856e-05, device='cuda:0', grad_fn=<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(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-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(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(7.9146e-05, 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(6.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4207e-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(5.1138e-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(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0338e-05, device='cuda:0', grad_fn=<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(0.0001, 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(0.0001, 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(9.3764e-05, device='cuda:0', grad_fn=<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(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<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(3.1822e-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(0.0006, 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(0.0003, 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(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9160e-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(0.0002, 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(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(8.6839e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2463e-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(0.0004, 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(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3328e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8163e-05, device='cuda:0', grad_fn=<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(8.4336e-05, device='cuda:0', grad_fn=<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(7.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4185e-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(0.0008, 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(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-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(0.0003, 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(6.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5962e-05, device='cuda:0', grad_fn=<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(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2968e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0468e-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>)
			------------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
				 {'precision': {'support': 0.8900662251655629, 'agreement': 0.5072463768115942, 'direct_attack': 0.1956521739130435, 'undercutter_attack': 0.2581967213114754, 'partial': 0.2962962962962963}, 'recall': {'support': 0.7804878048780488, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5206611570247934, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8316831683168316, 'agreement': 0.5660377358490567, 'direct_attack': 0.17307692307692307, 'undercutter_attack': 0.34520547945205476, 'partial': 0.2222222222222222}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6464497041420119, 'recall': 0.6464497041420119, 'f1': 0.6464497041420119, 'support': None}, 'macro_avg': {'precision': 0.4294915586995945, 'recall': 0.45486861118254956, 'f1': 0.42764510578341774, 'support': None}, 'weighted_avg': {'precision': 0.6879730383017154, 'recall': 0.6464497041420119, 'f1': 0.6588433303924994, 'support': None}}
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4711e-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(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(2.4616e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8990e-05, device='cuda:0', grad_fn=<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(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7163e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, 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(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1845e-05, 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(5.2894e-05, device='cuda:0', grad_fn=<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(3.8816e-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(0.0003, 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(6.1220e-05, 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(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(0.0002, 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(0.0004, 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(0.0006, 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(0.0006, 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(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1906e-05, device='cuda:0', grad_fn=<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(2.0135e-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(0.0003, 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(6.3308e-05, 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(0.0002, 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(2.6221e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9258e-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(2.5131e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-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(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3096e-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(3.6394e-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(8.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<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(0.0003, 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(0.0001, 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(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, 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(2.3466e-05, 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(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7647e-05, device='cuda:0', grad_fn=<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(5.5437e-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(7.4592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-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(0.0003, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4046e-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(9.6113e-05, 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(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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0603e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4692e-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(8.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2559e-05, device='cuda:0', grad_fn=<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(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, 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(4.5658e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9176e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9810e-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(7.9477e-05, 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(8.1959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7765e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0133e-05, device='cuda:0', grad_fn=<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(0.0004, 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(7.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2663e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<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(2.1589e-05, 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(3.2004e-05, 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(0.0011, 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(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0267e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, 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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2746e-05, 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(0.0004, 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(0.0003, 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(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5187e-05, 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(0.0001, 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(2.6009e-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(6.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2332e-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(0.0002, 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(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-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(0.0003, 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(2.5646e-05, device='cuda:0', grad_fn=<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(2.4616e-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(0.0007, 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(0.0001, 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(6.6701e-05, device='cuda:0', grad_fn=<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(8.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6413e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(6.6003e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, 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(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<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(3.2973e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9353e-05, 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(7.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8592e-05, device='cuda:0', grad_fn=<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(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8841e-05, device='cuda:0', grad_fn=<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(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(3.3245e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7177e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3716e-05, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, 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(7.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0432e-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(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(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_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(0.0002, 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(0.0003, 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(8.3866e-05, 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(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5108e-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(9.4620e-05, device='cuda:0', grad_fn=<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(2.8613e-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(8.8728e-05, device='cuda:0', grad_fn=<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(4.8823e-05, device='cuda:0', grad_fn=<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(8.4043e-05, device='cuda:0', grad_fn=<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(6.3248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8241e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-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(6.8273e-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(1.9378e-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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4258e-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(5.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(4.3358e-05, device='cuda:0', grad_fn=<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(5.1896e-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(1.8621e-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(2.2648e-05, device='cuda:0', grad_fn=<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(9.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(8.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6549e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-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(0.0003, 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(0.0002, 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(0.0004, 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(9.8220e-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(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3351e-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(2.3799e-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(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(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-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(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-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(0.0003, 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(8.0150e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<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(6.6880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5731e-05, device='cuda:0', grad_fn=<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(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2245e-05, 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(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, 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(5.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-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(0.0003, 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(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(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-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(0.0004, 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(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(4.2026e-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(5.3743e-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(4.0300e-05, device='cuda:0', grad_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(8.1047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-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(0.0005, 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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4539e-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(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, 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(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(6.4276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-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(7.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8805e-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(0.0003, 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(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7016e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6063e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8424e-05, device='cuda:0', grad_fn=<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(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, 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>)
			------------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
				 {'precision': {'support': 0.8900662251655629, 'agreement': 0.5072463768115942, 'direct_attack': 0.1956521739130435, 'undercutter_attack': 0.2581967213114754, 'partial': 0.2962962962962963}, 'recall': {'support': 0.7804878048780488, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5206611570247934, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8316831683168316, 'agreement': 0.5660377358490567, 'direct_attack': 0.17307692307692307, 'undercutter_attack': 0.34520547945205476, 'partial': 0.2222222222222222}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6464497041420119, 'recall': 0.6464497041420119, 'f1': 0.6464497041420119, 'support': None}, 'macro_avg': {'precision': 0.4294915586995945, 'recall': 0.45486861118254956, 'f1': 0.42764510578341774, 'support': None}, 'weighted_avg': {'precision': 0.6879730383017154, 'recall': 0.6464497041420119, 'f1': 0.6588433303924994, 'support': None}}
Loss: tensor(4.6022e-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(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(8.3113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3150e-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(4.1178e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3478e-05, device='cuda:0', grad_fn=<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(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, 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(9.4914e-05, 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(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(4.4144e-05, device='cuda:0', grad_fn=<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(7.8932e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<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(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3732e-05, device='cuda:0', grad_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(5.4347e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3598e-05, device='cuda:0', grad_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(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7638e-05, 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(2.4132e-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(7.5641e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8053e-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(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(3.7060e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-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(0.0002, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9192e-05, 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(8.9148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-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(0.0001, 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(9.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-05, 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(0.0003, 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(7.9943e-05, device='cuda:0', grad_fn=<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(9.0794e-05, 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(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9494e-05, 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(0.0002, 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(7.7356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7590e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6470e-05, device='cuda:0', grad_fn=<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(7.7135e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8163e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-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(2.4647e-05, device='cuda:0', grad_fn=<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(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-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(2.6403e-05, 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(8.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5312e-05, 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(6.4708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<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(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-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(0.0002, 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(1.8591e-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(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, 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(0.0002, 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(1.7743e-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(0.0003, 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(6.7064e-05, device='cuda:0', grad_fn=<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(2.2769e-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(0.0001, 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(2.2648e-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(2.6917e-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(9.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9462e-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(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6202e-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(8.9047e-05, device='cuda:0', grad_fn=<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(5.9479e-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(0.0004, 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(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(7.4444e-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(7.7632e-05, 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(2.1649e-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(0.0001, 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(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5712e-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(0.0002, 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(1.7350e-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(0.0004, 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(3.1005e-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(0.0001, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2243e-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(7.5572e-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(7.6256e-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(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(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(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6708e-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(1.4746e-05, device='cuda:0', grad_fn=<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(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7797e-05, 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(0.0001, 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(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, 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(3.5758e-05, device='cuda:0', grad_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(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7043e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4483e-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(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(9.6037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_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(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<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(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-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(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1331e-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(0.0006, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-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(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-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(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1746e-05, device='cuda:0', grad_fn=<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(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(0.0001, 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(2.5827e-05, 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(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4002e-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(2.5161e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8468e-05, device='cuda:0', grad_fn=<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(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2097e-05, device='cuda:0', grad_fn=<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(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, 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(2.7008e-05, device='cuda:0', grad_fn=<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(9.4535e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7550e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, 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(4.2994e-05, 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(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(7.3743e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-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(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(8.8164e-05, device='cuda:0', grad_fn=<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(2.1104e-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(6.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7320e-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(3.6576e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7428e-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(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, 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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2160e-05, device='cuda:0', grad_fn=<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(6.4440e-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(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-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(0.0002, device='cuda:0', grad_fn=<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(7.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-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(2.5525e-05, 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(0.0003, 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(6.3188e-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(2.0620e-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(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(7.3992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-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(6.9938e-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(9.2974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3905e-05, device='cuda:0', grad_fn=<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(0.0013, 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(7.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<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(8.3574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-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(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1638e-05, 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(7.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, 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(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6878e-05, device='cuda:0', grad_fn=<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(5.1587e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2585e-05, device='cuda:0', grad_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(0.0001, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-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(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6020e-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(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(0.0008, 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(7.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2690e-05, device='cuda:0', grad_fn=<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(6.0585e-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(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(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(7.7847e-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(0.0002, 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(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2431e-05, device='cuda:0', grad_fn=<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(7.2076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-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(9.4929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-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(0.0001, 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(0.0004, 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(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, 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(0.0003, 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(0.0013, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7487e-05, 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(7.8130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3634e-05, device='cuda:0', grad_fn=<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(2.6645e-05, 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(0.0003, 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(7.3038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<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(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(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3956e-05, device='cuda: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 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
				 {'precision': {'support': 0.8902116402116402, 'agreement': 0.5048076923076923, 'direct_attack': 0.1978021978021978, 'undercutter_attack': 0.2581967213114754, 'partial': 0.3018867924528302}, 'recall': {'support': 0.7816492450638792, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.5206611570247934, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8324056895485467, 'agreement': 0.564516129032258, 'direct_attack': 0.17391304347826086, 'undercutter_attack': 0.34520547945205476, 'partial': 0.22377622377622378}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.647189349112426, 'recall': 0.647189349112426, 'f1': 0.647189349112426, 'support': None}, 'macro_avg': {'precision': 0.4305810088171672, 'recall': 0.4551008992197156, 'f1': 0.4279633130574688, 'support': None}, 'weighted_avg': {'precision': 0.6883264447523535, 'recall': 0.647189349112426, 'f1': 0.6592940680914394, 'support': None}}
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<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(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-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(7.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4605e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2246e-05, 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(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, 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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-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(7.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<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(6.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, 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(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<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(3.2246e-05, device='cuda:0', grad_fn=<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(2.4586e-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(0.0001, device='cuda:0', grad_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(7.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-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(8.9645e-05, device='cuda:0', grad_fn=<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(0.0007, 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(9.2314e-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(0.0002, 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(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9700e-05, device='cuda:0', grad_fn=<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(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2814e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3201e-05, device='cuda:0', grad_fn=<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(4.3237e-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(0.0005, 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(8.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<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(8.8849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5835e-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(0.0003, 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(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, 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(0.0010, 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(0.0001, device='cuda:0', grad_fn=<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(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-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(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0309e-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(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, 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(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0541e-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(8.2867e-05, device='cuda:0', grad_fn=<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(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8819e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, 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(7.1179e-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(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-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(0.0004, 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(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, 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(6.1122e-05, 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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7217e-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(5.7867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-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(6.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-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(4.3660e-05, device='cuda:0', grad_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(0.0003, 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(8.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8966e-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(0.0003, 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(1.9923e-05, device='cuda:0', grad_fn=<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(2.2285e-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(5.0230e-05, device='cuda:0', grad_fn=<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(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-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(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3569e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6190e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-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(4.0421e-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(5.2107e-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(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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-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(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6011e-05, 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(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-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(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7968e-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(6.5263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3156e-05, device='cuda:0', grad_fn=<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(6.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5025e-05, 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(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1609e-05, device='cuda:0', grad_fn=<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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6743e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4529e-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(4.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9691e-05, device='cuda:0', grad_fn=<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(9.8204e-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(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<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(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-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(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9764e-05, device='cuda:0', grad_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(0.0002, 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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-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(7.7434e-05, device='cuda:0', grad_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(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, 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(1.8924e-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(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8324e-05, device='cuda:0', grad_fn=<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(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, 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(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, 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(2.8492e-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(1.8409e-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(0.0001, device='cuda:0', grad_fn=<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(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3108e-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(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(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7058e-05, 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(9.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3778e-05, 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(2.7372e-05, 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(2.5464e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7181e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7199e-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(0.0001, 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(6.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5439e-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(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0694e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7551e-05, device='cuda:0', grad_fn=<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(2.4041e-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(6.3489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9819e-05, device='cuda:0', grad_fn=<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(4.8309e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7450e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0829e-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(0.0002, 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(1.8470e-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(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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_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(9.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6954e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0513e-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(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1194e-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(7.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8525e-05, device='cuda:0', grad_fn=<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(6.6601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7674e-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(2.2679e-05, device='cuda:0', grad_fn=<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(0.0002, 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(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-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(0.0004, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-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(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1186e-05, device='cuda:0', grad_fn=<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(7.2075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8757e-05, device='cuda:0', grad_fn=<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(8.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3291e-05, device='cuda:0', grad_fn=<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(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7738e-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(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-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(7.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9975e-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(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6871e-05, device='cuda:0', grad_fn=<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(8.6516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9378e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3207e-05, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5241e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203e-05, device='cuda:0', grad_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-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(9.3140e-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(5.8070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9391e-05, device='cuda:0', grad_fn=<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(7.0513e-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(0.0003, 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(4.0658e-05, 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(0.0001, 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(5.5693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3171e-05, device='cuda:0', grad_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(1.6684e-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(2.2043e-05, device='cuda:0', grad_fn=<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(2.5797e-05, 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(5.8572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-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(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1472e-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(0.0003, 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(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9586e-05, 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(6.1639e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-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(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3326e-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(9.2695e-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(0.0002, 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(6.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5085e-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(6.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-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(4.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6413e-05, 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(4.6354e-05, device='cuda:0', grad_fn=<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(0.0001, 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(6.7758e-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>)
			------------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
				 {'precision': {'support': 0.8891820580474934, 'agreement': 0.5048076923076923, 'direct_attack': 0.1978021978021978, 'undercutter_attack': 0.256198347107438, 'partial': 0.3018867924528302}, 'recall': {'support': 0.7828106852497096, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.512396694214876, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8326127239036443, 'agreement': 0.564516129032258, 'direct_attack': 0.17391304347826086, 'undercutter_attack': 0.3415977961432507, 'partial': 0.22377622377622378}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.647189349112426, 'recall': 0.647189349112426, 'f1': 0.647189349112426, 'support': None}, 'macro_avg': {'precision': 0.42997541754353036, 'recall': 0.45368029469489823, 'f1': 0.4272831832667275, 'support': None}, 'weighted_avg': {'precision': 0.6874919229165407, 'recall': 0.647189349112426, 'f1': 0.6591030376915679, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, 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(5.9290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8667e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(5.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-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(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5717e-05, 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(1.4473e-05, 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(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9579e-05, 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(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<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(2.2467e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-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(9.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, 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(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, 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(8.4551e-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(6.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, 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(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3860e-05, 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(9.4481e-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(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3533e-05, 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(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8263e-05, device='cuda:0', grad_fn=<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(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-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(3.4456e-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(2.1891e-05, device='cuda:0', grad_fn=<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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1674e-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(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7091e-05, 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(0.0002, 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(4.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<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(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9673e-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(0.0001, 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(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9067e-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(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1973e-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(0.0002, 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(6.6042e-05, 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(2.1528e-05, device='cuda:0', grad_fn=<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(7.0073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8333e-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(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0071e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6296e-05, device='cuda:0', grad_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(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2883e-05, 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(3.6420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, 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(7.4779e-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(0.0002, 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(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8987e-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(9.6449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-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(0.0002, 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(2.1044e-05, device='cuda:0', grad_fn=<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(4.7505e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-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(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3660e-05, device='cuda:0', grad_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(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3117e-05, device='cuda:0', grad_fn=<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(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4054e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, 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(0.0003, 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(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7429e-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(9.1475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(3.9798e-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(6.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4018e-05, device='cuda:0', grad_fn=<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(7.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8853e-05, 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(5.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2120e-05, 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(1.7259e-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(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2552e-05, device='cuda:0', grad_fn=<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(7.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9753e-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(0.0003, 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(2.2467e-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(4.7202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-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(4.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, 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(9.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5800e-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(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(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(5.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0547e-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(7.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0001, 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(0.0001, 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(5.5420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<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(5.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4105e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0672e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7019e-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(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2299e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, 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(5.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-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(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-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(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(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(7.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-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(5.1221e-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(1.7925e-05, 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(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-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(2.7856e-05, 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(7.8648e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-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(0.0001, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5597e-05, device='cuda:0', grad_fn=<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(8.7099e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, 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(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7817e-05, device='cuda:0', grad_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(2.1891e-05, device='cuda:0', grad_fn=<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(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2998e-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(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8874e-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(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(9.9834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0347e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3587e-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(8.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-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(1.9711e-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(1.6714e-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(7.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9307e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7840e-05, 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(8.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8316e-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(4.7379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-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(0.0003, 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(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-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(6.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7290e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4077e-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(9.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4792e-05, device='cuda:0', grad_fn=<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(1.4473e-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(7.9778e-05, device='cuda:0', grad_fn=<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(5.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7703e-05, 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(6.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-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(5.6836e-05, device='cuda:0', grad_fn=<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(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6996e-05, device='cuda:0', grad_fn=<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(4.3600e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1580e-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(7.8421e-05, 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(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6347e-05, device='cuda:0', grad_fn=<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(2.0680e-05, device='cuda:0', grad_fn=<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(7.0359e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5444e-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(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3143e-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(7.4321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-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(8.6024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-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
				 {'precision': {'support': 0.8891820580474934, 'agreement': 0.5048076923076923, 'direct_attack': 0.1978021978021978, 'undercutter_attack': 0.256198347107438, 'partial': 0.3018867924528302}, 'recall': {'support': 0.7828106852497096, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.512396694214876, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8326127239036443, 'agreement': 0.564516129032258, 'direct_attack': 0.17391304347826086, 'undercutter_attack': 0.3415977961432507, 'partial': 0.22377622377622378}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.647189349112426, 'recall': 0.647189349112426, 'f1': 0.647189349112426, 'support': None}, 'macro_avg': {'precision': 0.42997541754353036, 'recall': 0.45368029469489823, 'f1': 0.4272831832667275, 'support': None}, 'weighted_avg': {'precision': 0.6874919229165407, 'recall': 0.647189349112426, 'f1': 0.6591030376915679, 'support': None}}
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0328e-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(5.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3695e-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(2.2406e-05, 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(5.2844e-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(9.1622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5262e-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(2.5343e-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(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4845e-05, device='cuda:0', grad_fn=<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(3.0218e-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(6.5194e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-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(8.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7414e-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(0.0002, device='cuda:0', grad_fn=<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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7726e-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(4.6481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-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(6.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1443e-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(8.2891e-05, 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(7.8006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7194e-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(6.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8658e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3123e-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(0.0001, 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(3.5546e-05, 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(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3131e-05, device='cuda:0', grad_fn=<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(9.0426e-05, 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(0.0003, 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(5.9713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0182e-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(0.0004, 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(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(6.1949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6271e-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(7.5652e-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(6.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1641e-05, device='cuda:0', grad_fn=<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(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7834e-05, 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(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5103e-05, 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(4.3183e-05, device='cuda:0', grad_fn=<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(2.3163e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-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(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4508e-05, device='cuda:0', grad_fn=<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(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5624e-05, 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(8.5767e-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(4.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6633e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-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(0.0002, 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(7.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8683e-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(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-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(0.0002, 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(9.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1127e-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(4.8358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2278e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, 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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6939e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-05, 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(8.7097e-05, device='cuda:0', grad_fn=<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(8.7320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4691e-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(0.0001, 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(4.0313e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-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(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(1.7592e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3245e-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(7.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0941e-05, 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(1.6290e-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(8.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9586e-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(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-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(1.5382e-05, 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(2.2588e-05, 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(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6623e-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(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2083e-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(4.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9125e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9369e-05, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1305e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5627e-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(5.9895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1673e-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(4.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-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(2.5252e-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(8.5657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_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(1.9378e-05, device='cuda:0', grad_fn=<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(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5619e-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(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-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(0.0002, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2908e-05, device='cuda:0', grad_fn=<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(6.6590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-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(6.6612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4501e-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(5.3469e-05, 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(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4278e-05, 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(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2251e-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(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, 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(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4423e-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(0.0002, 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(0.0001, 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(7.4596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9368e-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(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, 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(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1528e-05, device='cuda:0', grad_fn=<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(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-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(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6338e-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(0.0001, 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(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7028e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-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(9.4464e-05, 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(0.0001, 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(7.3389e-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(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8023e-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(5.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4413e-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(7.7391e-05, device='cuda:0', grad_fn=<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(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4200e-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(3.4658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, 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(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6819e-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(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-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(8.1867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1485e-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(6.8388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2724e-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(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(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-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(0.0003, 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(4.7621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0087e-05, 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(6.4629e-05, 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(3.9323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9849e-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(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, 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(8.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, 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(7.6450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-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(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, 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(8.9002e-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(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6063e-05, 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(6.5452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6899e-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(5.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0972e-05, 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(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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7384e-05, 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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-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(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5596e-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(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_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(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7592e-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(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-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(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, 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(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-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
				 {'precision': {'support': 0.8894736842105263, 'agreement': 0.5048076923076923, 'direct_attack': 0.1978021978021978, 'undercutter_attack': 0.25833333333333336, 'partial': 0.3018867924528302}, 'recall': {'support': 0.7851335656213705, 'agreement': 0.6402439024390244, 'direct_attack': 0.15517241379310345, 'undercutter_attack': 0.512396694214876, 'partial': 0.17777777777777778}, 'f1': {'support': 0.8340530536705737, 'agreement': 0.564516129032258, 'direct_attack': 0.17391304347826086, 'undercutter_attack': 0.34349030470914127, 'partial': 0.22377622377622378}, 'support': {'support': 861, 'agreement': 164, 'direct_attack': 116, 'undercutter_attack': 121, 'partial': 90}, 'micro_avg': {'precision': 0.6486686390532544, 'recall': 0.6486686390532544, 'f1': 0.6486686390532544, 'support': None}, 'macro_avg': {'precision': 0.430460740021316, 'recall': 0.4541448707692304, 'f1': 0.4279497509332916, 'support': None}, 'weighted_avg': {'precision': 0.6878687154163222, 'recall': 0.6486686390532544, 'f1': 0.6601896630360937, 'support': None}}
